Strasbourg Revisited: c'est chic de parler français

36
DOI ./ijsl-- IJSL ; : Penelope Gardner-Chloros Strasbourg revisited: c’est chic de parler français Abstract: This article describes the results of a follow-up sociolinguistic study to assess changes which have taken place in Strasbourg, France, since the early 1980s, when the situation there was studied by the same author. The linguistic context and the results of the original study are described, as well as the implica- tions and diculties of such real-time studies. Although the latest large-scale sta- tistical surveys are not directly comparable with the earlier ones, an accelerating shi away from the Alsatian dialect to the use of French is nonetheless evident. This results partly from diverse social changes aecting the city as opposed to the countryside. Alongside this, there is a snowball eect of trends which were already visible in the 1980s, reecting the loss of an elderly dialect-speaking gen- eration who helped maintain Alsatian within the family. Rapid anonymous sur- veys carried out in town show the dialect to be used only in specic “pockets”. Some six hours of conversations involving code-switching were also compared with a similar corpus collected earlier, and revealed a decline in more complex grammatical forms of switching in favor of more single-word switches. Alterna- tion was the dominant pattern although dierent proportions of switch-types oc- curred in dierent conversations. The functions of code-switching, e.g., for hu- mor and solidarity, remained similar. The conclusion reiterates the diculties but also the advantages of real-time sociolinguistic comparisons, as dierent socio- economic and lifestyle factors interact with ongoing linguistic shi and change. Keywords: code-switching, Alsatian dialect, French, age-grading, language shi Penelope Gardner-Chloros: Birkbeck, University of London. E-mail: [email protected] Introduction The slogan in the title of this article, which could best be translated nowadays as ‘It’s cool to speak French’, was put up in public places aer World War II to per- suade Alsatians to use French rather than their Germanic dialect, Alsatian. Some seventy years later, this message has largely been driven home, most especially in &6 :'* îPP '*0HWD6FLHQFH - ,-6/ SS ± ,-6/BB S 308:6/ 6HSWHPEHU $0

Transcript of Strasbourg Revisited: c'est chic de parler français

DOI 10.1515/ijsl-2013-0059   IJSL 2013; 224: 143 – 177

Penelope Gardner-ChlorosStrasbourg revisited: c’est chic de parler françaisAbstract: This article describes the results of a follow-up sociolinguistic study to assess changes which have taken place in Strasbourg, France, since the early 1980s, when the situation there was studied by the same author. The linguistic context and the results of the original study are described, as well as the implica-tions and di!culties of such real-time studies. Although the latest large-scale sta-tistical surveys are not directly comparable with the earlier ones, an accelerating shi" away from the Alsatian dialect to the use of French is nonetheless evident. This results partly from diverse social changes a#ecting the city as opposed to the countryside. Alongside this, there is a snowball e#ect of trends which were already visible in the 1980s, re%ecting the loss of an elderly dialect-speaking gen-eration who helped maintain Alsatian within the family. Rapid anonymous sur-veys carried out in town show the dialect to be used only in speci&c “pockets”. Some six hours of conversations involving code-switching were also compared with a similar corpus collected earlier, and revealed a decline in more complex grammatical forms of switching in favor of more single-word switches. Alterna-tion was the dominant pattern although di#erent proportions of switch-types oc-curred in di#erent conversations. The functions of code-switching, e.g., for hu-mor and solidarity, remained similar. The conclusion reiterates the di!culties but also the advantages of real-time sociolinguistic comparisons, as di#erent socio-economic and lifestyle factors interact with ongoing linguistic shi" and change.

Keywords: code-switching, Alsatian dialect, French, age-grading, language shi"

Penelope Gardner-Chloros: Birkbeck, University of London. E-mail: [email protected]

1 IntroductionThe slogan in the title of this article, which could best be translated nowadays as ‘It’s cool to speak French’, was put up in public places a"er World War II to per-suade Alsatians to use French rather than their Germanic dialect, Alsatian. Some seventy years later, this message has largely been driven home, most especially in

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

144   Penelope Gardner-Chloros

the capital, Strasbourg. The purpose of this study1 was to update a description of the linguistic situation in Strasbourg carried out in the 1980s (Gardner-Chloros 1985a, 1985b, 1991); as such it is one of a very small number of sociolinguistic comparisons made over such an extended period of time. The opportunity which this provides to view linguistic changes within a broader context of social change makes the exercise worthwhile, despite the di!culties in recreating the parame-ters of the original study exactly. For as the philosopher Heraclitus pointed out, you cannot step in the same river twice. This is at least as true in sociolinguistics as in other areas, and revisiting a linguistic situation turns out to be a more com-plex task than it would &rst appear.

More than 25 years had gone by since the original study, and there was little hope of &nding the same informants, or even necessarily the same settings for surveys. The aim was nevertheless to carry out a similar exercise to the original one on a smaller scale (there was only one month in which to collect data instead of the seven years spent there originally), and to make the closest possible com-parison with the information collected then. Code-switching was again to be used as a diagnostic tool to throw light on the speech habits of Strasbourgeois, in con-junction with an overview of statistical information available from various sur-veys, and a series of informal interviews with linguists and ordinary people. As before, conversational data was gathered in professional and in private settings and small-scale anonymous surveys in public places were carried out. It was hy-pothesized, based on a reading of published material which had appeared in the interim (OLCA 2012; Bister-Broosen 2002; Huck et al 2005), was that there would be much less dialect use in Strasbourg, perhaps to the point of non-existence in public settings and among the young. This hypothesis was indeed con&rmed. But the changes interacted with other developments which made giving a simple “be-fore and a"er” picture less than straightforward. There had been a shi" in atti-tudes and in ambiance with regard to linguistic matters, new factors, especially in the educational sphere, had become prominent, and any simple comparison with the earlier period appeared problematic. The river had moved on.

1 This research was carried out thanks to a Small Research Grant from the British Academy, whose support is gratefully acknowledged, and with additional support from the School of Social Sciences, History Philosophy at Birkbeck. A large number of people lent their assistance with the &eldwork, with helping me to understand the evolving situation, and with the analysis of surveys and recordings. I extend warm thanks to Valerie Bapst, Arlette Bothorel-Witz, Jenny Cheshire, Jean-Francois Clerc, Kevin Donnelly, Pascale Erhart, Piers Gardner, Corinne Gober-ville, Christine Helot, Dominique Huck, Francoise Hurstel, Benedicte Keck, Pierre Kretz, Anita Marchal, Jeremy McBride, Maria Secova, Andree Tabouret-Keller and Tatiana Termacic. Christine Fronsaq and Francois Leclerc deserve particular thanks for their assistance in collecting and transcribing recordings, as does Julia Hofweber for her help with the analysis.

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   145

Consequently it was necessary to focus not only on changes which had oc-curred in the use of languages as such, but also on changes in attitude, policy, urbanization and the broader cultural context. One was led to ask what could justi&ably be compared, rather than simply how things had changed. Comparing Strasbourg then and now was in some respects more like a diatopic comparison than a diachronic one – as the saying goes, the past is indeed “another country”, where things are done di#erently. This led to a consideration of how change has been treated in sociolinguistics. The study of linguistic change has been one of the linchpins of sociolinguistics since its earliest days (Labov 1963). But could the &ndings to do with real time, apparent time and age-grading studies be applied to change outside the variationist domain, in which the object of study is essentially con&ned to speci&c (phonological) variables? Or does a broader study like this one dictate a qualitative approach, illuminated by a few statistical pointers?

In the intervening period, sociolinguistics itself has also developed in numer-ous ways, many of which a#ect our view of minority languages. It would have been a surprise in the early 80s to be told that thirty years later, there would be more, and not fewer, linguistic minorities; yet thanks to continuing migration and globalization, this is undoubtedly the case.2 Languages are seen less and less as being connected to stable communities (Blommaert 2010; Heller 2006; Weber and Horner 2012); population movement is ever-increasing and language use changes accordingly. This leads not only to changing practices among members of these new communities, but also to new symbolic and virtual connections be-ing established, notably through the internet. Within this “superdiverse” world (Blommaert and Rampton 2011), linguists focus increasingly on how individuals construct their identities from disparate sources and resources rather than basing them on traditional certainties; they also try to understand how such identities are “brought about” within discourse itself, rather than being passively absorbed or inherited (Llamas and Watt 2010). Since the 1980s, there has also been a surge of interest in language endangerment and the possibilities for revitalization (Crystal 2000; Fishman 2001; Nettle and Romaine 2002). The emphasis within these publications is more on indigenous languages on other continents than on regional minorities in Europe such as Alsace, although Fishman (2001) includes chapters on Irish, Frisian, Basque and Catalan. As within France itself, minority languages which can be classi&ed – rightly or wrongly – as majority or o!cial languages of a neighboring state, have a bad habit of receiving less attention from linguists as from politicians, Gal (1979) being a notable exception.

2 According to the International Organization for Migration, in the last ten years alone, the total number of international migrants increased from approximately 150 million in 2000 to 214 mil-lion in 2008 (IOM 2008).

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

146   Penelope Gardner-Chloros

First, the relevance to the present study of diachronic comparisons which have been carried out in sociolinguistics is considered, bearing in mind the very di#erent focus and methodology of variationist analyses compared with the largely ethnographic description of a bilingual situation given here. Next, an overview of the current linguistic situation in Strasbourg is provided. Thirdly, the quantitative and qualitative information collected during a stay in the city in late 2011 is described. Section 4 contains an analysis of code-switching in a small cor-pus of everyday conversations, which is compared with a similar exercise con-ducted in the 80s. Naturally, this exploration of the use of Alsatian in Strasbourg only represents one small part of what could be studied in relation to the linguis-tic situation in the city. As in other urban centers, one could study the linguistic e#ects of migration from other areas and other countries, the emergence of “youth languages”, and, stemming more directly from dialectal issues, developments in the regional variety of French (Pellat 2010), as well as the evidence of “germanici-zation” in the dialect’s grammar (Bothorel and Huck 1998). All of these aspects have gained prominence since the original study. The fact of sticking here to the theme of dialect use and code-switching is a re%ection of time and space con-straints rather than the relative signi&cance of these changes. In the conclusion, I return to the question of how to study change in such contexts.

2 Studying sociolinguistic change

In his study of the diphthongs (ay) and (aw) in Martha’s Vineyard, Labov (1963) considered several explanations for the di#erences between younger and older speakers, and concluded that the best explanation rested on an “apparent-time” interpretation of the data. This interpretation suggests that the speech of older subjects re%ected earlier norms; therefore the correlations observed between age and the pronunciation of these vowels re%ected change in progress. An alterna-tive explanation, “age-grading”, also appeared to have some explanatory power, meaning that speakers also changed their pronunciation as they grew older. Most subsequent studies con&rmed the usefulness of the apparent-time interpretation (Chambers 2008; Sanko# and Wagner 2006); but several also con&rmed the exis-tence of age-grading e#ects. Sanko# and Laberge (1978) showed speakers in mid-adulthood making more use of standard variants, and Sanko# (2006) remarks that ultimately, the only way to decide which type of explanation is most informa-tive is to re-sample the community a"er a number of years, i.e. to carry out what she refers to a “trend study”. A French example can be found in the studies car-ried out by Ashby (1981, 2001) on the loss of the negative particle ne in spoken

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   147

French in Tours. By comparing the apparent time interpretation of the earlier study with a real-time follow-up, Ashby was able to show an acceleration in the rate of loss of the particle. He combined these methods with a “panel” study, meaning that the same individuals were followed up from the &rst study. Studies combining di#erent methods are unusual; they tend to show that only a minority of speakers change substantially over time. Harrington, Palethorpe and Watson’s (2000) panel study of Queen Elizabeth II’s vowels over 50 years of Christmas broadcasts showed her to be among those whose pronunciation did alter measur-ably over time. Overall, Sanko# argues that apparent-time studies which show a gradient age distribution are su!cient to identify change in progress.

The centrality of these issues for sociolinguistics, has led to a recent increase in research in which the apparent-time hypothesis is put to the test (Tagliamonte and D’Arcy 2004; Sanko# and Wagner 2006). Earlier real time frequently threw up new facts as well as con&rming the tendencies found originally. Trudgill (1988) found two new changes had entered Norwich speech when he replicated his 1974 study. Ashby (2001) found women’s relative rate of ne dropping changed once the feature had become more widespread overall. The recent studies tend to con&rm the results of Sanko# and Laberge (1978), in that language change and age- grading are seen to operate simultaneously. The use of non-standard variants tends to decline post adolescence, but some of them nevertheless spread when viewed diachronically.

The question is how linguistic change can best be studied when dealing with di#erent types of sociolinguistic change. While individual variables which lend themselves to repeated measurement over a large population, Dorian (2010) showed that language change in communities of di#erent size and di#erent types of social organization have to be approached in a di#erent way. In large, ethni-cally diverse and socioeconomically strati&ed cities, where most studies of varia-tion have been carried out, the variation correlates with ethnicity and class. Stras-bourg, with a population of 271,708 (INSEE 2009), may rank as a large urban centre, but it is made up of smaller communities, only some of which have long-standing local ties and speak the local dialect; this can even take on the status of a de&ning characteristic. Though to a lesser extent than in the very small com-munities investigated by Dorian, a high level of variation in dialect use results, here also, from the lack of extra-community norms for the minority language. German in particular is not seen as a pan-regional reference variety (Huck et al. 2005).

Within these smaller groupings, the choice between real and apparent-time investigation and the question of age-grading still have to be addressed. If we wish to know whether, and to what extent, a bilingual community is shi"ing towards monolingual usage of one of the varieties, then essentially the same options are

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

148   Penelope Gardner-Chloros

open as with the phonological studies: (1) we can synchronically compare the relative usage of the varieties by di#erent age-groups (apparent-time); (2) we can observe the age-grading within the population, preferably through “panel” studies in which the same individuals are followed over time; or (3) we can re-sample the community a number of years later. Option (1) is widely practiced, in the sense that most descriptions of minority language situations include tables such as that below, adapted from the results of a questionnaire sent out to house-holds in Alsace in 1979 enquiring about “Lifestyle in Alsace”. Table 1 is based on around 3,000 completed questionnaires.

The results, over 30 years ago, already indicated a stark falling o# in use (or knowledge) of the dialect. Unfortunately the question asked: “Do you speak the Alsatian dialect?” did not permit a clear distinction to be made between compe-tence and use. Secondly, one would wish to know the e#ect of a noticeably lower percentage of the younger age groups having been born in Alsace (83% of 16–24 year olds as opposed to the high nineties in the oldest groups). Thirdly, it is note-worthy that a slightly higher percentage of 16–24 year olds claimed to understand and speak the dialect than of 25–34 year olds, and similarly that 25% of the latter age group claimed not to understand or speak it at all, compared with 21% of the lowest age group. Since the survey was carried out in 1979, the 25–34 year old

Table 1: Distribution by age of replies to the question: “Do you speak the Alsatian dialect?”

Age

16–24 25–34 35–44 45–54 55–64 65–74 75 and over

(Average)

% understanding and speaking the dialect

66 64 71 84 84 88 88 78

% understanding but not speaking the dialect

13 11  9  5  5  4  2  7

% neither understanding nor speaking the dialect

21 25 20 11 11  8 10 15

(% of respondents born in Alsace)

(83) (90) (93) (96) (97) (98) (96) (93)

Source: Seligmann (1979)

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   149

group represents those born immediately a"er World War II, when there was a very marked reaction against all things Germanic, including the dialect, which many parents deliberately failed to pass on to their children. Later on, there was a reaction in turn against giving up the regional dialect, which explains the slight reprise in the youngest group. If the same questions had been asked in sub-sequent surveys, reliable trends might have been discernible, but unfortunately the survey was never repeated in the same form. More recent surveys are pre-sented below, but this example illustrates that even within one relatively simple set of statistics, an explanation based on gradient age distribution raises many questions, especially when there is a lack of linear progression. An age-graded progression is far from inconceivable, within which the youngest or next young-est groups might speak the dialect more and more as they grow older. As in the Sanko# and Laberge (1978) study, but particularly so for white-collar workers whose work environment would have demanded the regular use of the o!cial language, French, it could be that these speakers would use the dialect much more a"er retiring – providing they had learnt it to a su!cient extent in the &rst place.

Similarly, the analysis of code-switching in families and work places which was carried out in the earlier study showed the frequency of switching to be deter-mined to a considerable extent by setting, with more switching taking place in the workplace than in the homes. It is unlikely that the speakers in the work contexts were more balanced bilinguals than those recorded in their homes, because in both cases fairly balanced bilinguals were selected – though the presence of children in some of the home settings may have brought about more use of French. It seemed likely that the higher rate of switching at work was connected with the type of conversations which arose there as opposed to in the home. Those conver-sations involved a greater number of speakers, of varying competence in the dia-lect. Switching was therefore o"en used as an inclusiveness strategy, signaling solidarity among all the members of the group, including those with minimal dia-lect competence, and allowing them to participate fully in the conversation thanks to an abundant use of French. This type of motivation seemed to be over and above other motivations which were common to the family settings also, which included accommodation at an inter-individual level rather than the set-ting up of a kind of common (mixed) code as such. The earlier study is discussed here merely to point out some of the complexities of studying language distribu-tion and shi", compared with the variationist approach. Neither type of study is inherently more di!cult, or intrinsically less or more reliable. In fact, many of the same problems of interpretation arise, and in both cases factors other than pure statistical tendencies must be addressed. In both cases, the factor of age needs to be distinguished from the passage of time in order to explain linguistic change.

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

150   Penelope Gardner-Chloros

3  The linguistic situation in Strasbourg then and now

Di#erent aspects of the linguistic situation in Alsace overall are described in vari-ous publications and surveys, including Bister-Broosen (2002), Bothorel (2008), Denis et al. (1989), Erhart (2013), Duee (2002), Levy (2004), OLCA (2012), Selig-mann (1979), and Tabouret-Keller and Luckel (1981a, 1981b). Huck et al. (2005) provides a thorough discussion of linguistic issues which preoccupy speakers today (“les representations”), based on some 300 interviews.

Historically speaking, the dialects spoken in Eastern France can be traced back to invasions in the 4th/5th century by Germanic tribes, the Franks and the Alemanni. The written language in these areas was almost exclusively Latin until around the 13th century, but various dialectal forms of German were gradually written down from the 9th century onwards, and Luther’s translation of the Bible into the vernacular in the 16th century was a signi&cant step for the German lan-guage. From 1648, however, Alsace was attached to the French Crown. It was thoroughly gallicized following the French Revolution, and the advance of French was only halted when Alsace became German again at the end of the Franco-Prussian War (1870).3 It was returned to France a"er World War I, but was again annexed by Germany for &ve years during World War II. Very few speakers who were educated before 1918 are now alive, but many who were at school in 1940–1945 have memories of the part of their education which was in German. These frequent changes led to what has been termed the Alsatian “identity crisis” (Ho#et 1973; Phillips 1978). One of the features of this “crisis” is the speakers’ ambivalent attitude to their own dialect. A"er World War II, the teaching of Ger-man was prohibited at school for several years and young parents made a point of only speaking French to their children. As a result, the post-War generation started what may now be an irreversible decline in the use of the dialect, which was helped along by immigration from other areas of France and the imposed dominance of French in administrative, educational, public and media contexts. An INSEE survey in 1999, described in Duee (2002), involving 30,000 respon-dents, concluded that 500,000 adults (or 39% of the population of Alsace) spoke the dialect. This can be compared with Table 1, which shows that exactly 20 years

3  870–164 Alsace in Germanic Holy Roman Empire1648–1870 A"er 30 Years War, Alsace becomes French 1870–1918 A"er Franco-Prussian War, Alsace becomes German 1918–1940 A"er World War I, Alsace becomes French 1940–1945 Alsace annexed by Nazi Germany 1945– Alsace returned to France

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   151

earlier, 78% claimed to speak and understand it. Even more remarkable than this precipitous decline was the age gradient shown in the (1999) survey: only 10% of children born in Alsace were learning Alsatian, compared with 80% in the 1940s. Despite all this, and despite a certain amount of lip-service paid to the number of – linguistically and geographically disparate – Occitan-speakers (Duee 2002), this makes Alsatian still by far the most widely spoken regional language in France. See Figure 1.

Fig. 1: Percentage of dialect speakers (source: OLCA 2012)4

Throughout this period of decline, the dialect continued to enjoy consider-able covert prestige, and to be used for both informal communication, as well as literary and ludic purposes. In 1972 an enlightened school inspector, Holderith, introduced a teaching method in which the dialect was used to support the teach-ing of German in schools, and in 1975 the “Loi Haby” allowed optional courses in “Regional languages and Cultures” to be organized. Since 1990 there has been a growing movement to set up bilingual classes in French and German in schools, and by September 2011 there were some 20,000 pupils in primary schools in Alsace taking part in this successful scheme (Academie de Strasbourg, p.c.),

4  The ORBI and ISERCO surveys are local marketing surveys and all &gures are derived from OLCA (2012).

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

152   Penelope Gardner-Chloros

though the weekly contact hours in German have recently been reduced from 12 to 8. Alsatian itself is still not properly recognized as a language, and some mem-bers of the teaching profession – and even some linguists – argue that German is the only viable written form of the Alsatian dialects. There is no evidence, how-ever, that teaching German preserves the dialect among the young (Joseph 2007).

In 1993 an association of teachers wishing to encourage bilingual education in French and German was set up (the “Lehrer” Association), followed by a public organization to support and encourage the use of Alsatian in 1994, at the initia-tive of the regional government and the two administrative “départements” of the Upper and Lower Rhine, which make up Alsace. The functions of OLCA (“O!ce for the Language and Culture of Alsace”) include documenting as well as promot-ing the use of the dialect; thanks to this, more information about the dialect is being disseminated than was the case when reliance had to be placed on cen-suses by the national statistical o!ce INSEE (Table 1), where the linguistic infor-mation was incidental to the whole and the question ambiguously framed.

The results of this survey come 33 years a"er the &rst INSEE one, and provide food for thought with regard to the relative role of apparent time and age-grading. Although the exact circumstances of the two surveys are not the same – the INSEE one was a written questionnaire and the OLCA one was a telephone survey, both involved large and representative samples (INSEE: N = 3000; OLCA: N = 801, ad-justed to represent the factors of age, gender, profession and size of commune).

Fig. 2: Self-reported dialect pro-ciency in 2012 (source: OLCA 2012)

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   153

The average number from all age groups in the earlier survey who claimed to understand and speak the dialect was 88%; Figure 2 shows an absolute maxi-mum of 58% being in the same position now. Looking in more detail at the age groups, the 45–54 year-old group from the earlier survey can be compared with the over 60s in Figure 3. This would appear to indicate that there has been a change in the linguistic behavior of the same type of people: only 74% of over 60s claim to speak the dialect now, whereas in 1979, 84% of 45–54 year olds made the same claim. A similar tendency can be observed in the younger age group, those who were 16–24 in 1979 and who now fall within the 45–59 year old age bracket in Figure 3. Whereas 66% of them claimed to speak and understand the dialect then, only 54% do so now.

Can we therefore conclude that this age group has changed its linguistic be-havior over the intervening period? Since this age group, middle-aged at the time and now elderly, would have used mainly Alsatian with their own parents’ gen-eration, who have now mainly passed way, the change in behavior is to be ex-pected. Therefore the age-grading interpretation does appear necessary to under-standing of language shi" in such a context. Ideally, a number of further questions would need clarifying, for example, are we dealing with the same proportion of respondents born in Alsace in both surveys, or is there a higher proportion of non-locals in the more recent one? What is the role of rural-urban mobility? There are more town dwellers now than then, but the increase is in the suburbs rather

Fig. 3: Percentage of dialect speakers in di.erent age groups (source: OLCA 2012)

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

154   Penelope Gardner-Chloros

than the centre; this too may a#ect the statistics. Since Alsatian has long been spo-ken largely within the family, the passing of a generation has for a consequence quite simply that there are fewer Alsatian-speaking interlocutors around in most households; there are also fewer three-generation households overall. In the 1980s many young people said they spoke Alsatian primarily with their grandpar-ents; those grandparents are no more, and transmission within the family has therefore tailed o#. An age-grading e#ect, whereby young speakers might speak Alsatian more later on in life, can no longer realistically be expected in the light of the minimal competence observed in the youngest age groups.

This study focuses on Strasbourg. One problem with studying the city as op-posed to the region, then as now, is that there is a lack of systematic data on the di#erences between the two, though there are clear indications that the dialect is spoken much less in the capital than in rural areas. Northern Alsace, for example, is traditionally more dialect-speaking and also has a number of cross-border com-muters whose employment brings them into contact with German-speakers. But nearby villages which were also once dialect-dominant, such as Hoerdt, have moved towards becoming dormitory villages where the agricultural way of life has died out. Statistics provided by the OLCA survey, while not singling out Stras-bourg as such, provide some help towards clarifying these di#erences (Figure 4). In Strasbourg, with its large international population attached to the European institutions, and with the highest rate of in-migration from other countries and other parts of France in Alsace, the percentage of Alsatian speakers is more likely to be under 10% than the 20% reported for large municipalities in general. In the

Fig. 4: Percentage of dialect speakers depending on municipality size (source: OLCA 2012)

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   155

sections which follow, the nature of language use among the remaining hard core of dialect-speaking city dwellers is evaluated thanks to a mixed methodological approach.

4 Department stores and beyond

4.1 Printemps

In Gardner-Chloros (1985b), a variation on Labov’s famous study of postvocalic ‘r’ in department stores was described, focusing on language choices in Strasbourg. This exercise showed that rapid, anonymous surveys used by variationists to record the use of phonological features by large numbers of people could be adapted to provide information on language choices in bilingual situations. Such surveys were carried out in three Strasbourg department stores of di#ering social status, Printemps, Magmod and Jung, following the original New York study in Saks, Macy’s and Klein. Both salespersons’ and customers’ language choices were observed and noted down, each observation being classi&ed according to whether it was in Alsatian, French, or Alsatian and French combined. Variables considered, apart from the social class implicit in the hierarchy of department stores, included the estimated age of the speakers, whether they were in a depart-ment selling necessities or luxuries, and whether they were engaged in ingroup or outgroup interaction (i.e. salespersons among themselves v. salespersons to cus-tomers). Only female speakers were observed as they constituted the vast major-ity in both groups. As expected, more French was spoken higher up the social scale; and more Alsatian in ingroup rather than outgroup interaction, and by older speakers. The results as regards the code-switching category were less pre-dictable and could be ascribed to several motivations. There was a high rate of code-switching among young shoppers (the most francophone) in the most Alsatian-speaking store, but also among the oldest shoppers (the most dialect-speaking) in the most French-speaking store. This appeared to be due to compro-mise or accommodation to the linguistic environment. But other motivations were also re%ected in the distribution observed: for example the customer’s variety was more o"en switched to than the salesperson’s variety, re%ecting the power relations in this social situation; and overall speakers switched more when addressing the middle age group (30–45) than the older and younger groups, probably due to uncertainty about their preferences. The study highlighted the complexity of motives for language choices and especially for code-switching.

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

156   Penelope Gardner-Chloros

A perfect reproduction of the earlier study could not be achieved. Only Print-emps, the city-centre store highest up the social scale, still existed in the same form. The second department store, Magmod, had changed hands – and social character in the process – and the third one, Jung, had closed.

The same rapid anonymous study as before was therefore carried out in Print-emps only. Given the decline in dialect use outlined above, combined with the fact that Printemps gave rise to least dialect use even back in the 1980s, it was relatively surprising to &nd any use of Alsatian at all. Out of all the conversations observed, 90.4% were entirely in French, 7.7% involved code-switching and 1.9% were in Alsatian (N = 104). This compares with the earlier study, in which 81.8% of conversations observed were in French, 10.2% involved code-switching and 8% were Alsatian (n = 88) (Gardner-Chloros 1991: 74). Although the rate of dialect use has gone down markedly, there remains a “hard core” of speakers who continue to use it even in this cosmopolitan, city-centre commercial environment where the linguistic landscape – names of products and shop signs of all types – is reso-lutely French. Figure 5 shows the breakdown by age: unsurprisingly, the use of Alsatian and switching is still skewed towards the oldest age group. Figure 6 shows the breakdown in relation to ingroup/outgroup behavior, and reveals that by far the most Alsatian – and code-switching – is found among customers. By contrast, in the 1980s, salespeople among themselves used 13.6% Alsatian and 18.2% switching (Gardner-Chloros 1991: 80). A likely explanation for this change is that whereas a fair proportion of customers may be well over retirement age, salespeople are by de&nition younger.

Fig. 5: Language choices in Printemps according to age group

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   157

4.2 Markets

In order to compensate for the lack of the other two department stores, two studies were undertaken in Strasbourg’s open air food markets, which have had a very long-standing presence in the city. Markets are generally fruitful locations to ob-serve linguistic choices and the vernacular in particular (e.g. Petrits 1990). They bring rural speakers (the stall-holders and farmers) into contact with urban dwellers, many of whom are regular customers and develop routines and spaces of repeated interaction. On the basis of advice by several locals, it appeared that if Alsatian was making a showing in any public setting in Strasbourg, the markets – especially outside the center – would be the best locations to &nd it. Two of the largest, the food markets of Boulevard de la Marne and the suburb of Neuhof, were both extended by further rows of stalls selling clothes and other items, held largely by stall-holders of immigrant (largely North-African and Turkish) origin. In this instance only the food sections were observed, but the ethnic sections would doubtless provide fruitful material for a further study of a quite di#er-ent range of instances of language contact. The age pro&le of food market shop-pers is skewed in favor of older age-groups, so in order to get the clearest possible picture of the dialect’s survival in the city and not to overestimate the number of

Fig. 6: Language choices in Printemps according to salesperson/customer status

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

158   Penelope Gardner-Chloros

Alsatian speakers, a decision was taken to observe only those who appeared to be up to 40 years old. Speakers of obviously non-European ethnicity were also omitted.

Figure 7 presents the data obtained in terms of the proportion of each lan-guage used in each of the two locations. Given the relatively small number of conversations involving Alsatian in the under 40 age group, those where the dia-lect was used on its own and those involving both varieties were combined into one category. The results show that Alsatian is more commonly used in the sub-urban context (12.1%, N = 157) than in the centre (4.2 %, N = 187). The overall aver-age is 7.9% for the two markets. If the over 40s’ age group had been observed it is fairly safe to assume that this &gure could have been at least trebled.

In this case a division according to gender was possible, but as Figure 8 shows, although men used slightly more Alsatian than women, the di#erence was unremarkable.

Finally, as in the Printemps study, the language choices of sellers were com-pared with those of customers. Sellers, perhaps because many were not actually residents of Strasbourg but came in once a week for the market, used a consider-ably higher proportion of Alsatian compared with customers (ingroup and out-group interaction were not distinguished). See Figure 9.

Fig. 7: Language choices among under 40s at Boulevard de la Marne and Neuhof markets

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   159

To sum up regarding the rapid anonymous surveys, the information obtained con&rms that the use of the dialect in Strasbourg is highly circumscribed. In the case of Printemps, the only context which could actually be replicated from the previous study, a hard core of some 10% of speakers continued to use at least some Alsatian, down from approximately 18% in the early 1980s. This 10% was to

Fig. 8: Language use by gender

Fig. 9: Language use broken down by customer/salesperson

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

160   Penelope Gardner-Chloros

be found overwhelmingly in the older age groups, and therefore, owing to the comparative demographic pro&le of customers and salespeople, among customers rather than salespeople. The two marketplace surveys were not directly com-parable with the previous study. Over 40 year olds, who would certainly have in-%ated the &gures for Alsatian uage, were not observed, in order to gauge the ex-tent of dialect use among younger speakers and provide more information about likely trends in the future. The more suburban of the two markets, Neuhof, pro-vided a &gure of over 12% Alsatian usage; the more central one well under this &gure. There appeared not to be any remarkable gender di#erences, but there was a di#erence between sellers and buyers, with sellers using the dialect more, prob-ably due to their rural provenance.

One may conclude that it is still possible to speak Alsatian in Strasbourg in certain “safe havens” and delimited circumstances, but that these have become notably more circumscribed over the last 30 or so years. The dialect is used es-sentially with interlocutors one knows to be dialect speakers as well, or in spe-ci&c contexts where it is clear that its use is “permissible”. But it is becoming a much less categorical choice even among the oldest speakers, the latter generally being balanced bilinguals rather than dialect-dominant as the oldest speakers were in 1985. It is arguable that dialect use now is more determined by choice, and less determined by competence, than it was in the 1980s. The next section will address the question of how the two varieties are organised in the locals’ speech. This will be done through an analysis of a corpus of conversations involving both languages from professional and private settings, again replicating the procedure of the earlier study.

5  Analysis of conversations involving code-switching

5.1 Introduction

Code-switching is one of the most sensitive barometers of language use in bi-lingual situations, with di#erent types of switching characterizing bilingual com-munities at di#erent stages in their evolution and the distribution of varieties within code-switching re%ecting those varieties’ respective sociolinguistic role and vitality (Deuchar et al. 2007; Gardner-Chloros 2009; Muysken 2000).

At a practical level, just as the settings for the anonymous surveys required could not be reproduced perfectly from the earlier study, so the process of obtain-ing natural recordings from urban dwellers who used the dialect in their everyday

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   161

lives also proved something of a challenge. In the end, it proved impossible to &nd young speakers who used the dialect regularly – though one cannot con-clude that they do not exist at all. For example two young mothers in their twenties who spoke the dialect to one another at work in Strasbourg, and to their children at home, were interviewed, but both lived outside the city itself and their children went to school in the nearby villages where they lived, so they could not be included in the study. The following recordings were obtained, four in profes-sional settings and four in domestic/social settings.5 See Table 2.

The total length of recordings was just over 5h and 52 minutes, some 4 hours and 30 minutes in domestic/social settings, and approximately 1 hour and 22 minutes in professional settings. The professional setttings were all shops and therefore had more constrained interactions than the more varied settings recorded in 1985. The total number of words transcribed was 16,159 (11,861 in domestic/social settings and 4,298 in professional/commercial ones). There were 62 speakers in all (vs. 31 in 1985), this number being in%ated by the customers com-ing in to the shops brie%y to buy something and leaving again. None of the main speakers was under 40 and the average age of all speakers was 60. The average

5 All names given here are pseudonyms.

Table 2: Recordings

Domestic/social

Mme Huck 1 24min 43secMme Huck 2 6min 04secMonsieur Kramm 1 17min 21secMonsieur Kramm 2 27min 17secMme Hainault 17min 06secMonsieur Schlumpf 1 53min 21secMonsieur Schlumpf 2 64min 26secMonsieur Schlumpf 3 61min 55sec

Professional/commercial

Antique Shop 12min 02secButcher’s 07min 49secBaker 1 20min 18secBaker 2 15min 47secStationer’s 26min 27sec

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

162   Penelope Gardner-Chloros

age of speakers was not calculated in the earlier study but was approximately 38, and the social mix was similar. This means that the recordings probably do not di#er markedly from what would have been obtained had it been possible to re-record the same speakers 25 years on. The Huck, Kramm and Schlumpf conver-sations were in several parts due to interruptions, but the interlocutors and the situation were unchanged, so they have been treated as a single conversation in each case. Non-transcribed material included inaudible or extended mono lingual passages in either French or Alsatian: for example Monsieur Schlumpf’s conver-sation was largely in French, so only the passages where both languages were used have been exploited. The transcription and coding system used was CLAN (MacWhinney 2007), within which computational tools are available for extract-ing various combinations which are of interest.

In terms of socio-professional categories, the majority of speakers were middle-class, except Madame Huck who was working class (retired cleaner). Other professions include white collar o!ce workers, retired members of the mil-itary, and retired company executives. Direct correlations cannot easily be made between socio-professional category and language choice in Strasbourg. True working class people might be expected to speak more dialect, but the inter-section with the factors of age and rural/urban provenance – or indeed migrant origin – makes such an assertion problematic. Working in certain environments (e.g. commercial or administrative) which require contact with the broader public or with colleagues from other parts of France, whether in person or on the phone, encourages the use of French, which o"en then becomes habitual; this factor which may explain Mr Schlumpf’s extensive use of French.

5.2 Analysis

The categories of code-switching analyzed were based on the (1991) study, and as  then, an attempt was made to quantify them. They included single-word switches, divided up according to parts of speech; switches between independent clauses, between dependent clauses, within clauses (intra-clause) and “disjointed” or paratactic switches (e.g. those coinciding with le"-dislocation). Inter-turn switches were not included in the earlier study, but represent a signi&cant cate-gory here. “Expressions” includes exclamations, discourse markers, greetings, swearwords and the like. See Figure 10.

The frequency (FREQ) command available within CLAN could only reliably be used in cases where more or less the full recording was transcribed. It was therefore applicable to all the domestic/social settings except Mr Schlumpf, but

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   163

not to the professional/commercial settings. The COMBO command, which could theoretically have been used to count switches, had to be complemented by a manual count of all switches for technical reasons.6

Figure 11 shows the evolution in the proportion of switch types between 1985 and now, leaving aside the inter-turn switches. Inter-turn switching is interesting above all from a Conversational Analysis perspective, an analytical approach which was &rst applied to bilingual data some years a"er the original work done (Auer 1999; Li Wei 2005). Space considerations do not allow for a detailed analy-sis of the inter-turn switches along CA lines here, although some relevant com-ments are made below.

6 CLAN transcriptions cut turns into a series of short utterances (the utterance has a mixed grammatical and intonational de&nition, cf. Gardner-Chloros et al. [2000]; Gardner-Chloros et al. [2007]). This can mean that a dependent clause and a main clause may be transcribed as two di#erent sequential utterances by the same speaker. Switches at those points did not carry the “@s” code which identi&es code-switches, unlike the switches within utterances. Since CLAN’s “Combo” command only identi&es switches marked “@s”, switches at that point in the utterance were not picked up. This issue arose essentially in the domestic/private conversations, where there were longer passages of speech. In the conversations in shops, such as the Butcher’s, the CLAN Combo command could be used because the actual utterances were short and so corre-sponded well with the division into utterances. Thus di#erent techniques provided within the CLAN toolkit were appropriate for di#erent types of conversation. To compensate for this, each individual switch was identi&ed manually, as in the original study.

Fig. 10: Overall distribution of switch types

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

164   Penelope Gardner-Chloros

Fig. 11: Comparative table of types of switches in 19857 and in current study

The only type of switch which increased slightly over the period was single-word switches. Given that more balanced bilinguals tend to use more complex switching (Poplack et al. 1988; Rindler-Schjerve 1998: 247), it is possible that these bilinguals were less balanced than the previous cohort. This result is some-what surprising given that the speakers were from a similar social background as before and of the same age that the earlier ones would have been, on average, if they had been re-recorded. This may lead us to surmise an age-grading e#ect.

Further clues are provided by the direction in which speakers switched. Did they mainly start o# in French and change over to Alsatian, or vice versa? There are many di!culties with deciding – and even more with quantifying – whether speakers switch more from language A to language B or vice versa, which have to do with the unruly nature of spontaneous speech, interruptions, overlaps and inaudible sections. It is possible, however, to count insertions8 of individual word-switches in sentences from the other language. These excluded street

7 The recordings described in Gardner-Chloros (1991) were made around 1985.8 The idea that such words are “inserted” does not imply that a particular grammatical theory of code-switching is being espoused. In particular, the notion that there must always be a “base” or “matrix” language for each utterance does not underlie this classi&cation. The assumption that there is always a base language can be argued to beg the interesting questions rather than answering them. For a fuller discussion, see Gardner-Chloros (2009: Ch. 8) as well as Muysken (2000: 64–69).

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   165

names, greetings and politeness formulae such as s’il vous plait, which were con-sidered loans rather than code-switches, along with a small number of other well-established terms, such as pressing, the French term for ‘dry-cleaner’. As before, there were a few mixed words such as antiquitégschae!, antique shop.

As before also, the commonest category of single-word switches were French nouns in Alsatian utterances, and as before this is a re%ection of the fact that the dialect operates within a francophone context, particularly as regards Ausbau terms (e.g. les symphonies de Beethoven, funiculaire, compagnie). Tellingly how-ever, there was a higher proportion of Alsatian words in French utterances than in 1985. One recording, the antique shop, had an overwhelming proportion of in-sertions of French terms in Alsatian utterances and none going the other way, an untypical pattern which partly minimizes the extent of this change. Figures 11 and 12 should be read together. As the more complex grammatical switching de-clines, a more emblematic type of single-word switching appears to be taking its place.

Finally, a further analysis of the switches in these recordings was carried out based on Muysken’s (2000) in%uential division of switches into three categories: Insertion, Alternation and Congruent Lexicalization, based on their grammatical characteristics. Muysken argued that each of these three types of code-switching is typical of di#erent types of language contact situation (2000: 246). He placed

Fig. 12: Direction of single-word switching then and now

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

166   Penelope Gardner-Chloros

the three di#erent types of code-switching on the three corners of a triangle, indicating that there is a continuum between them. The languages and locations in the centre represent code-switching corpora discussed by di#erent research-ers. The Strasbourg data contained instances of all three types and so was to be found towards the middle of the triangle. See Figure 13.

The data from the current study shows alternation to be the dominant pattern in all but one recording. The antique shop recording is an “outlier” in that it con-sists in a largely Alsatian conversation with 58 single word switches into French – many to do with specialized “antique shop” vocabulary (e.g, dorure d’origine ‘original gilding’), a bare handful of other switches into French and no switches from French into Alsatian. Comments on congruent lexicalization will be found below. See Figure 14.

The conversations analyzed here have been placed in a similar triangle, in order to illustrate, in the same manner as Muysken, the variety of code-switching pro&les to be found within a single speech community; such internal variation has been relatively little examined in the literature so far. See Figure 15.

Fig. 13: Three types of code-switching (Muysken 2000: 246)

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   167

Fig. 15: Variation of predominant code-switching patterns

Fig. 14: Proportion of insertions, alternations and congruent lexicalizations in current data-set

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

168   Penelope Gardner-Chloros

The triangle shows alternation to be the dominant pattern (overall there was 67% alternation, 31% insertion and 2% congruent lexicalization), which accords with Muysken’s view that alternation occurs predominantly in stable bilingual com-munities with a tradition of language separation. At the same time, the triangle provides an indication of the variations to be found between the various recordings.

5.3 Qualitative remarks9

In the &nal section, instances of switching which occurred widely in these conver-sations are brie%y discussed and exempli&ed, both in terms of form and conver-sational functions. As we saw in Figure 11, essentially the same type of switches occurred in the earlier and later recordings, though in di#erent proportions. Single words, from discourse markers (doch! ‘but yes’) and exclamations (ah ben voila! ‘well there you are’) to mots justes (baptême de l’air ‘maiden %ight’; haupt-stross ‘main street’) and swearwords, o"en with a positive politeness function (dreckspatz (lit.) ‘dirty sparrow’), were the commonest sites for switching. As before, some of these consisted in French words phonologically and morphologi-cally adapted to Alsatian, o"en in preference to the German equivalents (e.g. elektrizität for ‘electricity’ instead of German STROM, personalität for ‘personal-ity’ instead of German PERSÖNLICHKEIT; decidiere for ‘to decide’ instead of German ENTSCHEIDEN). The fact that the regional variety borrows lexical items from French rather than the linguistically more closely related High German under lines the political nature of the relationship between standard language and dialects/minority languages. Greetings, terms of address, and common po-liteness terms were, as ever, in French (Bonjour, madame, merci) and acted to frame dialect utterances within a French context, a kind of direct linguistic trans-position of the social situation. As a grammatical pattern, alternation was domi-nant, with inter-turn switches, switches between independent clauses and “dis-jointed” or paratactic switches make up the majority of cases. As before, many switches coincided with the onset of subordinate clauses and with link words such as but, because, when, if, near, a!er, etc and relatives such as who, thus highlighting both the grammatical and semantic/discourse connections between the two parts of the sentence, for example:10

9 French: italics; Alsatian: plain font; German: CAPITALS; xxx: inaudible.10 The transcription conventions adopted here for Alsatian, as in the earlier study, are from Matzen (1980).

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   169

(1) Mich het als d’grossmamm usg’lacht parce qu’on habitait à côté du foyer des jeunes

‘Grandma always used to laugh at me because we lived next to the youth hostel’

In some cases, the link word itself was the only switched element, as in this example, which also shows typical paratactic qualities:

(2) Une fois c’est tôt, une fois c’est tard . . . tu n’as pas de . . . awwer . . . le truc de euh de lune . . . de lune, exact, et je me lève pour le plaisir

‘Sometimes it’s early, sometimes it’s late . . . there is no . . . but . . . the thing about the uh the moon, exactly, and I get up for the pleasure of it’

There were a small number of instances of congruent lexicalization, in which words from either language are integrated within a common grammatical frame, as there were in 1985. An example from this corpus, with a switch between pro-noun subject and verb which shows an unusual degree of convergence in the grammars, is:

(3) C’était qu’il stinkt ‘It (i.e. the problem) was that it stinks’

A comparable example from 1985 was the utterance:

(4) ça griwelt ‘it prickles’

Such instances, as said, were exceptional, which accords with Muysken’s com-ment that congruent lexicalization occurs either between closely related lan-guages, (e.g., in dialect-standard switching in Dutch or Italian) or where there is no tradition of overt language separation (e.g. second generation migrant groups, post-creole continua); the languages are assumed to partly share their processing systems. There is no evidence emerging from this work of the distinction between French and Alsatian becoming less marked at a sociolinguistic level.

Among the switch-types which were common to both data-sets were numer-ous switches to mark reported speech, for example:

(5) Ja (.) j’me suis dit wenn d’r ebbs passiert bisch nitt allaan c’était ça. ‘Yes (.) I told myself if something happens to you you’re not alone that

was it.’

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

170   Penelope Gardner-Chloros

Reiteration of what was said in the other language was also common, whether between speakers or within utterances. Some of these simply act as backchannels whereby speakers con&rm their understanding or being on the same wavelength as their interlocutor, for example, at the antique shop, when the dealer repeats the client’s reference to “dimension” ‘dimension’ as “grösse” ‘size’. Sometimes the repetition is somewhat more complex discourse-wise, as when a client at the stationer’s is complaining to the shopkeeper that she never wins at the lotto. Probably stung by a perceived implication that this is something to do with the tickets being sold there, the shopkeeper remarks that you have to buy more tickets in order to win, and does so &rst in one language, and then in the other, render-ing  the point more emphatic. Alternatively one can view the &rst comment in French as a neutral piece of information, and the repetition in Alsatian as a friendly tip.

(6) shopkeeper: vous jouez pas assez c’est pour ça qu’vous gagnez pas ! ‘you don’t play enough that’s why you don’t win!’ customer: xxxx . shopkeeper: spiele nitt g’nug! ‘don’t play enough!’

As before, parataxis, which can pass largely unnoticed in casual speech, was extensively exploited to avoid mixing the two grammars, for example:

(7) Même (. . .) bon, l’activité physique, het abg’numme, hein? ‘Even (. . .) well, physical activity, there’s less of it, eh?’

Parataxis is sometimes accompanied by other devices such as doubling, com-monly used to get round syntactic clashes in code-switching. In the following example, the inversion of verb and subject, which would be necessary in Alsatian in the subordinate clause, is avoided through the repetition of ‘a"er that’ (noch-här) in French (après ça):

(8) und no läs ich ung’fähr a halb Stund und nochhär (. . .) après ça j’essaye de m’endormir

‘and then I read for about half an hour and then (. . .) a!er that I try to fall asleep’

There were also some di#erences from the earlier study:– Fewer switches to mark changes of addressee: here almost all speakers were

pro&cient in both varieties, whereas in the earlier study the age range was

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   171

greater and people frequently switched to French to address children, or to Alsatian to address the elderly.

– Switches between speaker-turns were not discussed in 1991. On this occasion they were found to be numerous. Grammatically speaking, they occurred be-tween independent clauses, between dependent clauses and within clauses, as in the frequent case of interlocutors &nishing o# each other’s turns in the other language.

The following extract provides an example of this and of collaborative/cohesive language choices:

(9)Speaker 1: wenn ich net kann schlofe a virt’l stund nochhär dass i’s licht

usg’macht hab - ‘When I can’t sleep a quarter of an hour a"er turning out the light-’Speaker 2: Ta nuit, elle est foutue ‘Your night is wrecked’ Speaker 1: je n’ (. . .), en$n, le (.), le milieu de la nuit est foutu ‘I d- (. . .), well, the (.), the middle of the night is wrecked’

A “shared” switch between dependent clauses occurs between the &rst two turns, when Speaker 2 &nishes o# the Speaker 1’s utterance. Speaker 1 then starts o# a  new sentence, accepting Speaker 2’s language choice; a"er a false start he accepts Speaker 2’s interpretation as well, repeating his words with a slight quali&cation.11

Finally, the dialect’s function in humor and the creation of rapport is as lively as ever. As locals frequently point out, the savor of the dialect is an essential ele-ment in jokes and anecdotes among Alsatian-French bilinguals, especially where these jokes have an earthy side. For example, in a characteristic sequence, Mme Huck’s interlocutor asks her to tell the story – obviously a favorite – about the “de falsch Grumbeere” ‘the fake potatoes’ saying “Vous me ferez d’Grumbeere?” ‘Will you do (sic) me the potatoes?’ The story is about her when she was a child, asking her mother why she always had a big belly – her mother was in fact almost al-ways pregnant, a topic which was not deemed suitable to explain to such a young girl:

11 The word order of the if-clause “wenn ich net kann schlofe” is French rather than the Alsatian (“wenn ich net schlofe kann”). There is considerable variation – even in the same individuals – in which of these orders prevails in Alsatian clauses.

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

172   Penelope Gardner-Chloros

(10)Madame Huck “Mamma, ich versteh’s nitt ! Du bisch imm’r dick . . . un hasch

imm’r a so dick’r Buch” . . . “Ja mädl s’kummt vun d’Grumbeere . . .”

‘“Mummy, I don’t understand! You are always fat . . . and always have such a fat tummy” . . . “Yes dearie that’s because of the potatoes . . .”’

Speaker 2: Vum Grumbeere esse . . . ‘From eating potatoes’ Mme Huck Ja! Et ça c’était . . . ah bon! Bon, les patates, hein (laughter) ‘Yes! And it was. . . . Oh I see! Right, the potatoes, eh.’

The language of her childhood was Alsatian, but the adult joke can best be shared in – and structured thanks to – a code-switched mode. It is the humorous motive rather than historical exactness which is paramount in the language choice. The end of the story is marked by “Ja”, used as a discourse marker. She then switches to French to ironically paraphrase what went on in her mind at the time “Ah bon? Bon, les patates, hein!”, reporting her naïve amazement at the answer.

While the form of code-switching may change more or less subtly as the lan-guage balance shi"s, its associations and functions show remarkable longevity.

6 Conclusions

In the future, real-time follow-up studies may increasingly take the place of apparent-time studies thanks to the increasing availability of sound recordings of all types. For example, more and more speakers, especially those in the public eye, are captured repeatedly on YouTube: a recent study conducted in this way at Manchester University showed how the celebrity couple, David and Victoria Beckham, have altered their pronunciation over the last 15 years to sound more “posh” (http://www.manchester.ac.uk/aboutus/news/display/?id=9856, accessed 26 August 2013). As such studies become more numerous, it may become easier to choose between competing explanations for the changes observed. If age-grading can be shown to be signi&cant over as short a period as 15 years, then the role of accommodation and other socio-psychological factors in language change may need to be re-assessed.

This study showed some of the reasons why investigating change in minority language situations and monitoring linguistic vitality pose a number of complex challenges. The changing linguistic balance between a national standard and a

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   173

regional minority language is a function of multiple in%uences, some based on observable changes, such as increased in-migration and educational reforms; some more imponderable, such as evolving attitudes towards heritage languages; and some resulting from an implacable geometric progression, as older generations die o# and the availability of interlocutors in the minority language dwindles. Dorian’s (2010) study of the evolution of Gaelic in Scottish &shing villages, where linguistic variation was unrelated to social class and the other classic social groupings, provides an excellent demonstration of the complexity of the factors involved in studying language use. Di#erent results are obtained from communi-ties with di#erent forms of social organization, such as that of the remaining hard core of Alsatian speakers in Strasbourg. The fact that it was not possible, as origi-nally planned, to carry out a replication of the earlier Strasbourg study, results from changes in that social organization interacting with more speci&cally lin-guistic changes.

Nevertheless, some conclusions may be drawn about the manner in which, and the extent to which, language shi" has occurred in the intervening period. First of all, it can be con&rmed, perhaps surprisingly, that Strasbourg has not become a dialect-free zone. Use of the dialect has undoubtedly shrunk, as all the surveys show. Yet the city’s lack of homogeneity has helped the dialect to survive in certain pockets – for example in the old market garden village, turned suburb, of the Robertsau – and among certain speakers, who may hang on to it all the more tightly – and emotionally – as they sense that their numbers are shrinking. Larger-scale surveys and the anonymous surveys carried out in this study show a steep age-gradient which could presumably only be reversed if large numbers of  village-dwellers moved into town. Despite this, some revival of interest in Alsatian can be detected. Miss France 2012, who came from Alsace, gave an ac-ceptance speech in which she said she hoped she could be an ambassador for the language, and indeed uttered some words of Alsatian in the speech itself. This would have been unthinkable 25 years ago, when Alsatian still had more rustic and masculine connotations, despite being more widely spoken by both sexes. The increased exposure to German at school is a sign of the &nal attenuation of post-War anti-German sentiments, which also had their e#ect on the dialect. But while the proponents of such a bilingual education argue that German is the writ-ten form of the dialect, the Ausbau language remains French, and pupils’ learn-ing of German is unlikely to a#ect the long-term survival of the dialect. To this one must add an almost universal reluctance to contemplate standardizing Alsatian, which other minority language experiences suggest is tantamount to a death-wish.

The surveys served to frame an analysis of bilingual speech based on a cor-pus of conversations. In order to identify dialect speakers – the only ones who were in a position to code-switch, except in a purely symbolic way – and record

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

174   Penelope Gardner-Chloros

them within a reasonable space of time, the study concentrated on the over-60s age group. Younger speakers might have been found with more thorough search-ing, but they were certainly in a small minority. Since the original group of record-ings was made with socio-economically similar speakers with a much lower aver-age age, this exercise can be considered a real time study. Providing we accept that these older speakers were essentially the same kind of speakers as before, we can ask to what extent age-grading also played a role, i.e. in what way, if any, their code-switching had changed since 1985. In many – or most – ways, it had re-mained the same. But some pointers, such as the fact that there was a higher proportion of single word switches compared with multiple word ones than had been found before, suggests that some change in speakers’ code-switching habits may have occurred. The fewer people there are to share the language with, the less activated it is, and the more likely that it will survive in the form of single word switching – largely concept-tied – rather than in balanced code-switching.

The article began with the statement that it is “chic” to speak French, a piece of propaganda which there is no longer any necessity to push in Strasbourg. The large presence of foreigners, whether Eurocrats, other immigrants or “Français de l’Intérieur” – but also Alsatians who commute in from surrounding villages – today completes the picture of a singularly unhomogeneous patchwork, within which dialect-speakers feel both %attered at their capital’s cosmopolitanism and besieged in their regional identity. The decrease in dialect use is not only a matter of personal choice – where choice there is – but also a matter of changing life-styles and of the environment within which that choice is exercised. The last word may go to the refrain from one of the best-known songs of Germain Muller, the quintessential Alsatian poet and wit: « Mir sin schins d’letschte, ja d’allerletschte von denne laetze wo noch so babble wie de schnawwel ‘ne gewachse ésch » [We are, it seems, the last, yes the very last of those twisted types who still speak “the way their beak grew”].

ReferencesAshby, William J. 1981. The loss of the negative ‘ne’ in French: a syntactic change in progress.

Language 57. 674–687.Ashby, William. J. 2001. Un nouveau regard sur la chute du ne en français parlé tourangeau:

s’agit-il d’un changement en cours? Journal of French Language Studies 11. 1–22.Auer, Peter (ed.). 1999. Code-switching in conversation: language, interaction and identity.

London: Routledge.Bister-Broosen, Helga. 2002. Sprachkontakte und Sprachattitüden Jugendlicher im Elsaß und in

Baden: Vergleichende soziolinguistische Untersuchungen in Colmar (Frankreich) und in Freiburg und Müllheim (Deutschland). Frankfurt: Peter Lang.

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   175

Blommaert, Jan. 2010. The sociolinguistics of globalization. Cambridge: Cambridge University Press.

Blommaert, Jan & Ben Rampton. 2011. Language and superdiversity. Diversities 13( 2). UNESCO. ISSN 2079-6595. www.unesco.org/shs/diversities/vol13/issue2/art1 (accessed 26 August 2013).

Bothorel-Witz, Arlette & Dominique Huck. 2000. Die Dialekte im Elsaß zwischen Tradition und Modernität. In Dieter Stellmacher (ed.), Dialektologie zwischen Tradition und Neuansätzen, Beiträge der Internationalen Dialektologentagung Göttingen, 19–21 Oktober 1998 (Zeitschri0 für Deutsche Literatur Beihe0 109), 143–155. Stuttgart: Franz Steiner.

Bothorel, Arlette. 2008. Le plurilinguisme en Alsace : les représentations sociales comme ressources ou outils de la description sociolinguistique. In Recherches en didactique des langues – L’Alsace au coeur du plurilinguisme (Les Cahiers de l’ACEDLE 5[1]), 41–63. http://acedle.u-strasbg.fr (accessed 26 August 2013).

Chambers, Jack. 2008. Sociolinguistic theory. Oxford: Wiley-Blackwell.Crystal, David. 2000. Language death. Cambridge: Cambridge University Press.Denis, Marie-Noële , Calvin J. Veltman & Monique Wach . 1989. Le déclin du dialecte alsacien.

Strasbourg: Publications de l’Université de Strasbourg.Deuchar, Margaret Pieter Muysken & Sung-Lan Wang. 2007. Structured variation in code-

switching; towards an empirically based typology of bilingual speech patterns. International Journal of Bilingual Education and Bilingualism 10(3). 298–340.

Dorian, Nancy. 2010. Investigating variation: the e"ects of social organization and social setting. Oxford & New York: Oxford University Press.

Duee, Marc. 2002. L’alsacien, deuxième langue régionale de France. http://www.insee.fr/fr/insee_regions/alsace/themes/cpar12_1.pdf (accessed 26 August 2013).

Erhart, Pascale. 2013. Les dialectes dans les médias : quelle image de l’Alsace véhiculent-ils dans les émissions de la télévision régionale ? Strasbourg: Université René Descartes unpublished thesis.

Fishman, Joshua A. (ed.). 2001. Can threatened languages be saved? Reversing language shi#, revisited – a 21st century perspective. Clevedon: Multilingual Matters.

Gal, Susan. 1979. Language shi#: social determinants of linguistic change in bilingual Austria. New York: Academic Press.

Gardner-Chloros, Penelope. 1985a. Le code-switching à Strasbourg. In Gilbert L. Salmon (ed.), Le français en Alsace, 51–60. Paris: Champion-Slatkine.

Gardner-Chloros, Penelope. 1985b. Language selection and switching among Strasbourg shoppers. International Journal of the Sociology of Language 54. 117–135.

Gardner-Chloros, Penelope. 1991. Language selection and switching in Strasbourg. Oxford: Oxford University Press.

Gardner-Chloros, Penelope, Melissa Moyer & Mark Sebba Mark. 2000. The LIDES coding manual: a document for preparing and analyzing language interaction data. [Special issue]. International Journal of Bilingualism 4(2).

Gardner-Chloros, Penelope, Melissa Moyer & Mark Sebba. 2007. Coding and analysing multilingual data: the LIDES corpus. In J. C. Beal, K. P. Corrigan & H. Moisl (eds), Models and methods in the handling of unconventional digital corpora (Synchronic Corpora 1), 91–120. Houndmills, Basingstoke, New York: Palgrave Macmillan.

Gardner-Chloros, Penelope. 2009. Code-switching. Cambridge: Cambridge University Press.

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

176   Penelope Gardner-Chloros

Harrington, Jonathan, Sally-Anne Palethorpe & Catherine Watson. 2000. Monophthongal vowel changes in Received Pronunciation: an acoustic analysis of the Queen’s Christmas Broadcasts. Journal of the International Phonetic Association 30. 63–78.

Heller, Monica. 2006. Linguistic minorities and modernity: a sociolinguistic ethnography, 2nd edn. London, New York: Continuum.

Ho.et, Frederic. 1973. Psychanalyse de l’Alsace. Colmar: Alsatia Poche. Huck, Dominique, Arlette Bothorel-Witz & Anemone Geiger-Jaillet. 2005. Report on the

linguistic situation in Alsace (France). Language Bridges Papers. www.eurac.edu/languagebridges (accessed 30 May 2013).

INSEE. 2009. http://www.insee.fr/fr/bases-de-donnees/esl/comparateur.asp?codgeo=com-67482 (accessed 26 August 2013).

IOM (International Organization for Migration). 2008. United Nations Department of Economic and Social A"airs (UN DESA), trends in international migrant stock: the 2008 revision. http://esa.un.org/migration/index.asp?panel=1 (accessed 26 August 2013).

Joseph, John E. 2007. 842, 1871, and all that: Alsace-Lorraine and the transformations of linguistic nationalism. In Wendy Ayres-Bennett & Mari C. Jones (eds.), The French language and questions of identity, 44–52. London: Modern Humanities Research Association and Maney Publishing.

Labov, William. 1963. The social motivation of a sound change. Word 19. 273–309.Levy, Paul. 2004. Histoire linguistique d’Alsace et de Lorraine. Paris: Editions Manucius.Li Wei. 2005. “How can you tell?” Towards a common sense explanation of conversational

code-switching. Journal of Pragmatics 37(3). 375–389.Llamas, Carmen & Dominic Watt. 2010. Language and identities. Edinburgh: Edinburgh

University Press.MacWhinney, Brian. 2007. The TalkBank Project. In Joan Beal, Karen Corrigan & Hermann Moisl

(eds.), Creating and digitizing language corpora: synchronic databases, Vol. 1, 163–180. Houndmills, Basingstoke, New York: Palgrave-Macmillan.

Matzen, Raymond. 1980. Dichte isch bichte. Gedichte in Strassburger Mundart. Kehl: A. Morstadt.

Muysken, Pieter. 2000. Bilingual speech. Cambridge: Cambridge University Press.Nettle, Daniel & Suzanne Romaine. 2002. Vanishing voices: the extinction of the world’s

languages. Oxford: Oxford University Press.OLCA. 2012. Etude sur le dialecte alsacien http://www.olcalsace.org/sites/default/-les/

documents/etude_linguistique_olca_edinstitut.pdf (accessed 26 August 2013).Pellat, Jean-Claude. 2010. Les variétiés régionales de français en Alsace: représentations

et interprétations. In Dominique Huck (ed.), Parole(s) et langue(s), espaces et temps: mélanges o"erts à Arlette Bothorel-Witz, 203–211. Strasbourg: Presses de l’Université.

Petrits, Angeliki. 1990. Addressing in Modern Greek: Evidence from a case study in the Athens Central Market. Sociolinguistics 19. 125–144.

Phillips, Eugene. 1978. L’Alsace face à son destin: la crise d’identité. Strasbourg: Société d’Edition de la Basse Alsace.

Poplack, Shana, David Sanko. & Christopher Miller. 1988. The social correlates and linguistic processes of lexical borrowing and assimilation. Linguistics 26. 47–104.

Rindler-Schjerve, Rosita. 1998. Code-switching as an indicator for language shi0? Evidence from Sardinian-Italian bilingualism. In Rodolfo Jacobson (ed.), Codeswitching worldwide, 221–248. Berlin & New York: Mouton de Gruyter.

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

Strasbourg revisited   177

Sanko., Gillian. 2006. Age: Apparent time and real time. In Keith Brown (ed.), Encyclopedia of language and linguistics, 2nd edn, Article number LALI: 01479. Oxford: Elsevier.

Sanko., David & Susan Laberge. 1978. The linguistic market and the statistical explanation of variability. In David Sanko. (ed.), Linguistic variation: models and methods, 239–250. New York: Academic Press.

Sanko., Gillian & Susan Evans Wagner. 2006. Age grading in retrograde movement: the inflected future in Montréal French. University of Pennsylvania Working Papers in Linguistics 12(2). 1–14.

Seligmann, Nicole. 1979. Connaissance declarée du dialecte et de l’allemand. Chi"res pour l’Alsace 26(4). 21–30.

Tabouret-Keller, Andree & Frederic Luckel. 1981a. La dynamique sociale du changement linguistique: quelques aspects de la situation rurale en Alsace. International Journal of the Sociology of Language 29. 51–66.

Tabouret-Keller, Andree & Frederic Luckel. 1981b. Maintien de l’alsacien et adoption du français; éléments de la situation linguistique en milieu rural en Alsace. Languages 61. 39–62.

Tagliamonte, Sali & Alex D’Arcy. 2004. He’s like, she’s like: The quotative system in Canadian youth. Journal of Sociolinguistics 8(4). 493–514.

Trudgill, Peter. 1988. Norwich revisited: recent linguistic changes in an English urban dialect. English World-Wide 9. 33–49.

Weber, Jean-Jacques & Kristine Horner. 2012. Introducing multilingualism: A social approach. Abingdon, Oxon, New York: Routledge.

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940

123456789

10111213141516171819202122232425262728293031323334353637383940

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

�&6��� :'*�����î���PP�� '*0HWD6FLHQFH�� � -������,-6/������ SS�����±���� ,-6/B���B��������� �S������308��:6/������������� ���6HSWHPEHU������������$0

12345678910111213141516171819202122232425262728293031323334353637383940