The study

In this article, I explore two commonly occurring themes in interviewees’ accounts. These themes have been selected in relation to the problem outlined above. The role of intentionality and agency in intergenerational transmission points towards local-scale communicative patterns commonly examined in studies of family language policy. In most cases, of course, intergenerational transmission is what makes the difference between language maintenance and language shift (Fishman Reference Fishman1991:113). As the discussion shows below, however, the relationship between language maintenance and intergenerational language use is more complicated than oftentimes assumed, and it becomes clear that macro perspectives have to be involved in any study of the reasons behind language shift. In fact, even in more micro-oriented scholarship, there is evidence pointing at factors beyond the family unit (e.g. Caldas & Caron-Caldas Reference Caldas and Caron-Caldas2000; Canagarajah Reference Canagarajah2008).

Therefore, the following section examines the problem from the opposite view. Proceeding from more global-scale processes, the external pressures on the community to shift language is the focus here. The discussion is not, however, centered on the external pressures as presented by the last-speaker generation, but rather the first-generation nonspeaker viewpoint serves as the point of departure. Clearly, the adult generation in nontransmission would have a more extensive understanding of what the external pressures were (and, in the case of Gammalsvenskby, some of these have been probed in Forsman Reference Forsman2015) but, by looking at how the child generation remembers those external pressures, we can gain insights into the social factors involved in nonacquisition of a language which—according to the testimony of most of my informants, regardless of competence in the language—was regularly and extensively used during the time period in question.

The ‘linguistic culture’ (Schiffman Reference Schiffman and Ricento2006) of the community is traced through the overt statements of sixteen first-generation non-Swedish speakers in Gammalsvenskby (born between 1936 and 1970). The loosely structured interviews were made in the Gammalsvenskby area in August 2013 in Russian and focused on the reasons for the collective language shift, questions of ethnic identity, and village life in general. To an extent, the statements of these nonspeakers are contrasted with ideas expressed by members of the parent-generation in sociolinguistic and life-history interviews made in Swedish in 2001, 2011, and 2013.

The aim is thus to study the language shift in question bottom-up, proceeding from the accounts of single everyday situations and moving towards more general developments over time. Particular attention is given to the interviewees’ accounts of language use in the family, but these are related to more general issues in the life of the community and of the village in the decades during which the interviewees grew up.

In the text, the terms micro and macro are sometimes used, even though they have come under increasing criticism in contemporary sociolinguistics (see e.g. Blommaert Reference Blommaert2007). One reason for using these terms is that they occur in much of the work cited. Another is that they are a convenient way of referring to the kind of perspective taken on the analysed event in each instance. In this article, these terms are thus not to be understood as bounded, constant, or even discrete levels. Rather, macro refers to a top-down analytical perspective, whereas micro correspondingly refers to a bottom-up perspective. The methodological basis for this article is, however, that more global and more local discourses are interlinked and thus show different effects depending on the perspective taken.

A short overview of societal developments in Gammalsvenskby 1929–1950

As the text below makes reference to the global-scale historical events in Gammalsvenskby (which are, of course, also local-scale events in the reality of the individual) during the period of language shift among the Gammalsvenskby Swedes, this section gives a short background to those events (for more detailed accounts, see Hedman & Åhlander Reference Hedman and Åhlander2003 or Kotljarchuk Reference Kotljarchuk2012).

Gammalsvenskby is part of modern-day Zmiïvka, a rural community in Khersons'ka Oblast’ (Kherson Province) in southern Ukraine. As mentioned above, it seems that the year 1929 marks the onset of collective language shift among the Gammalsvenskby Swedes. In that year, most of the Swedish community in Gammalsvenskby (some 1,000 persons) emigrated to Sweden. Not all Swedes in the area emigrated—for various reasons—and, more importantly, in the following two years approximately one quarter of those who had emigrated returned to Gammalsvenskby. The reasons for the return seem to have been manifold, but among them, one important factor was that the Gammalsvenskby Swedes had been scattered over different parts of Sweden and they felt it was impossible to maintain their traditional, close-knit village culture and way of life under such circumstances. Most of the last speakers remaining today are descendants of returnees. It is clear that the numerical and proportional strength of the Swedish-speaking community decreased significantly as a result of the majority staying in Sweden, which in turn would have influenced how well the remaining Swedish speakers could assert their language on the local level (cf. Forsman Reference Forsman, Koskensalo, Smeds, de Cillia and Huguet2012). Furthermore, the returnees soon realised that they would not be able to keep in touch with friends and family in Sweden, since international correspondence came to be seen as potential foreign intelligence activities. Thus, the community was cut off from any community of Swedish speakers beyond the village itself, a reality further explored below.

In 1937, eighteen Swedish-ethnic villagers were detained and subsequently executed or tortured to death as part of the Stalin-era repressions. Although the repressions hit people all over the Soviet Union randomly on the whole, some of my last-speaker informants claim that there were also elements of ethnic cleansing in them. Around the same time, the early Soviet policy of korennizacija (often translated as nativisation or ‘putting down roots’), which, at least in principle, aimed at some degree of cultural autonomy for linguistic and ethnic minorities, was ended (Martin Reference Martin, Suny and Martin2001:73f; Vähä Reference Vähä, Chulos and Remy2002:105). In practice this meant, among other things, that the Swedish school in the village was closed down and that religion was abolished. This had linguistic consequences for the Gammalsvenskby Swedes, because the school and the church were the places where children learnt Standard Swedish, which was another way of tying them to a bigger ethnic and linguistic picture.

The 1940s involved great upheavals in Gammalsvenskby, as elsewhere in Europe. Between 1941 and 1943, much of Ukraine was occupied by the German army. As the village was part of a village complex with ethnic German majority, and many of the cultural practices of the ethnic Swedes were more or less similar to those of the local ethnic Germans, the Swedes were given status as Volksdeutsche, a sort of racial definition of ethnic Germans. This status gave the Swedes some privileges during the occupation, but it also led to their being forcibly evacuated to Germany when the German army retreated (Gaunt Reference Gaunt and Bobyleva2007). In Germany, most of them lived in work camps with other Volksdeutsche and prisoners of war, and were employed as forced labour in agriculture and the war industry.

At the end of the war, almost all Gammalsvenskby Swedes were ‘repatriated’ to the Soviet Union. Regarded as enemies of the people, the majority were sent to GULAG camps in the Komi republic. The conditions there were harsh and many, particularly children and the elderly, died from starvation, cold, or hard labour. Due to extraordinary circumstances, most of those still living were, however, allowed to leave the camps and return to their home area in 1947. As the village had been exposed to large-scale shelling in 1943, housing was scarce, and many settled in other parts of the village complex (selo) instead. Previously the village complex had been neatly divided into one Swedish and three German villages, but after the war there was no ethnic division between the villages.

The social structure of the village and community changed further in the beginning of the 1950s. First, there was a great influx of ethnic Ukrainians into the area, mainly because of the border adjustments between Poland and the Soviet Union. Theses Ukrainians stemmed from Western Ukraine, and were thus distinct both culturally and linguistically from those few minority Ukrainians that had previously lived in the area. Furthermore, a large dam was built at Kakhovka some 10 km down the Dnieper, resulting in the river covering parts of the village as well as valuable islands and fishing waters. This, together with more explicit Soviet-wide policies, permanently altered the socioeconomic structure of the village and the basis for its traditional culture.

FLP and endangered languages: Micro and macro in the study of language maintenance and shift

Despite early scholarship on language policy focussing on explicit and overt (e.g. state-wide) language planning, in recent years there has been a turn towards studying the more implicit ideologies that shape linguistic practices in everyday life (Spolsky Reference Spolsky2012). The concept of FLP implies that values, beliefs, and practices of family members (parents, siblings, grandparents) as well as other persons in one's immediate environment, shape linguistic practices at least as much as—often more than—the official policies and attitudes of society at large (Curdt-Christiansen Reference Curdt-Christiansen2009; Spolsky Reference Spolsky2012).

Within endangered-languages scholarship, it is, however, often emphasised that collective language-shift processes are tied not only to the micro-level linguistic behaviour of individuals, that is, linguistic transactions within the family or in the immediate, everyday context. Instead, such studies also recognise the effects of more or less global-scale discourses. Among such discourses are the linguistic values and policies of society at large. This includes both overt policies, such as attitudes towards multilingualism by the state or local society, and covert ones, such as interethnic relations and majority members’ views on the minority group. Other discourses, more close to the local level and often termed ‘meso’, are collective identities and language attitudes that prevail within the group, the local community, and so on. These, too, are sometimes overt—as in a community organising formal teaching in the community language—and sometimes covert—such as general pressure towards endogamy or attitudes towards code-switching (see e.g. Tandefelt Reference Tandefelt, Fase, Jaspaert and Kroon1992:151; Grenoble & Whaley Reference Grenoble, Whaley, Grenoble and Whaley1998; Musk Reference Musk2010).

In the case of the Gammalsvenskby language shift, the most common FLP methodologies are not directly applicable, since the language shift under scrutiny is one that occurred some sixty to seventy years ago. Therefore, it is not possible to make observations about home language, or longitudinal studies of the child generation's linguistic competencies. This is why the data from the interviews with members of the nonspeaker generation have to be seen in contrast with the results of language shift/language death studies building on the parent generation's accounts, including the parent-generation accounts from Gammalsvenskby.

However, ‘the family is not a self-contained institution that can adopt its own strategies and devices for language transmission’ (Canagarajah Reference Canagarajah2008:170). Endangered-languages scholars have often focussed on the more global-scale processes of language shift. This may partly be due to researchers’ ideologies, as many have wanted to point out that the community undergoing language shift cannot be blamed exclusively for the lack of intergenerational transmission, but that there are larger, societal forces operating in processes of language shift (cf. e.g. Dorian Reference Dorian1981; Edwards Reference Edwards1985). In recent years, there has also been much discussion about the ‘linguistic ecological catastrophe’, whereby it is implied not only that the loss of lesser-used languages is a process in which speakers of all languages are involved, but also that the preserving of linguistic diversity should be a concern for everyone, whether or not they are speakers of languages threatened with extinction (e.g. Mühlhäusler Reference Mühlhäusler1996:322–24; Skutnabb-Kangas Reference Skutnabb-Kangas, Skutnabb-Kangas, Phillipson, Mohanty and Panda2009:38f). These currents, however, often tend to overlook the micro perspective in language shift, namely the communicative reality of the speakers (cf. King, Fogle, & Logan-Terry Reference King, Fogle and Logan-Terry2008:909). On the one hand, it is certainly true that language shift in many, if not most, cases is an extension of sociocultural/linguistic policies and practices of dominant linguistic cultures, which govern the values, practices, and choices of individuals, filtered through the micro-reality in which they exist (Edwards Reference Edwards1985:51; Canagarajah Reference Canagarajah2008). On the other hand, however, that micro-reality being what it is, the difference between individuals who are participating in language shift and individuals who are not, is not only how those individuals are positioned in relation to dominant societal discourses, but also how they position themselves in relation to their respective realities (Edwards Reference Edwards1985, Reference Edwards1994). Therefore, a holistic study of language shift needs to take both the macro and the micro perspectives into account.

Opposing views with a common denominator: The role of agency in intergenerational transmission in Gammalsvenskby

A common notion in FLP scholarship is that intergenerational transmission of a language is oftentimes not particularly dependent on overt linguistic policies of the family, but rather on implicit socialisation to communicative patterns that prevail in the family (Schwartz Reference Schwartz2010; Fogle & King Reference Fogle and King2013). By contrast, scholars working on situations of language endangerment often focus on agency (e.g. Beck & Lam Reference Beck and Lam2008; Pérez Báez Reference Pérez Báez2013). In such studies, one may find mentions of parents choosing another language (or several others) over the heritage language for a variety of reasons, such as a wish to increase their children's academic or economic prospects, or making it easier for the children to adapt to other demands of the surrounding society (Tsunoda Reference Tsunoda2005:57–63 and works cited there). Whatever the reasons cited, it seems that the question of agency and intentionality in intergenerational language transmission is a central point in the study of language shift, whether collective or individual.

In the case of the Gammalsvenskby Swedes, it is interesting that many interviewees, both from the child generation and from the parent generation, report that Swedish was always spoken in the homes. One example from an interview with a non-Swedish speaker is given in (1).Footnote ¹

1. (1) I:  no u vas eeh doma gavarili pa-švedsk?

   ‘But did you speak Swedish at home?’

   F1960s1:  da → babuška jesl- eh kagda žyla iščo maja babuška. (.) pastajanna pa-švedski havarili

   ‘Yes, grandmother if- eh when my grandmother was still alive, [they/ we] spoke Swedish all the time.’

   I:  mhm

   F1960s1:  ja kagda šla u pervyj klas mama havarila što ja russkava jezyka ni znala

   ‘Myself, when I went to first grade, mother has told me that I didn't know any Russian [= Ukrainian].’

   I:  mhm

   F1960s1:  va abšče ni znala jak (laughs) ni znala kak ty budeš škole učitsja (.) t! eeh a tuda zabyla pašla u školu zabyla → nu::: ja ah pomnila ja da ja ja znaju kak havarjat pa starošvedskomu ja fsjo panimaju no: havarit’ ja ni havarila. a teper’ nas-nač-načinaju učit'sja.

   ‘[She] didn't know at all how (laughs) didn't know how you are going to study at school. Tsk! Eh, and there I forgot, went to school, forgot. Well, I remembered yes I, I know how they speak in the Gammalsvenskby dialect, I understand everything, but to speak, I didn't speak. But now I have started to learn.’

This interviewee, born in the late 1960s, claims that both parents and grandparents spoke only Swedish in her childhood home, to the extent that she did not know any Ukrainian at all before going to school. Yet today, although she seems to be a rather competent passive bilingual, her active skills in the home language are close to none. In adulthood she has been participating actively in evening classes in Standard Swedish, and in the extract it is noteworthy that she equates the Swedish of her evening classes with the local Swedish of her parents, although the two varieties are quite divergent and not automatically mutually intelligible. This ties well into the theme of a wider linguistic horizon discussed below.

The account in (1) is echoed in the accounts of many of the last speakers. In particular, those whose endogamous marriages resulted in non-Swedish-speaking children claim that Swedish was always spoken in the home, but that it was the pressures of school or, later in life, the children's exogamous partners that led to a shift in language choice between the parents and their children. The role of exogamy as a factor in language shift is often brought up (Tsunoda Reference Tsunoda2005:51–52 and sources cited there), but just a quick glance at the circumstances in Gammalsvenskby shows both that some of the non-Swedish-speaking interviewees grew up in endogamous families, and that several of the last speakers grew up in exogamous families. Here it may be noted that of this interviewee's two living siblings neither the older brother nor the younger one seems to have much active competence in the dialect. Although the family was not only endogamous (yet with the mother stemming from an exogamous family), but also seems to be one where a sense of Swedish-ethnic identity has been particularly actively upheld, the children lost the language despite reportedly active use at home in childhood.

Whereas the interviewee in (1) says that she only spoke Swedish in childhood, other interviewees report that they never spoke nor were expected to speak Swedish. One, only two years younger, even told me (on several occasions) that her grandmother explicitly told the children that they did not need Swedish, given in (2). She ascribes this conscious nontransmission to fear in the Soviet society.

1. (2) F1970s1:  možet byt’, ja na primer dumaju što eta vremja byla takoje. možet byt je- vot kagda v tritsatsidmom gadu zabrali ijo atsa. eeh byli pristalinskie represi. možet byt’ ljudi bajalis tagda ja dumaju. [patomušta-

   ‘Maybe, myself for instance, I think that the times were such. Maybe if- see, when they took her father away in -37. Eeh, it was the Stalin-era repressions. Maybe people were afraid back then, I think. Because-’

   I:   [no eta uže byla (.) bolše tritsati let na[zad?

   ‘But that was already more than thirty years back?’

   F1970s1: [aa znaješ kakuj strach kak glubakoj sedit (.) u::: ljudej ljudi bajatsa:: dolga patom.

   ‘Ah, you know what deep fear there is in people! People are afraid for a long time afterwards.’

   I: mmhm nu da.

   ‘Mm, well yes.’

   F1970s1:  patomušto ot ja pomnju ana nam gavarila @<eta> vam nji nada.@

   ‘Because, you see, I remember she told us “This, you don't need [it]”.’

   I:  mm

   F1970s1:  ano vam nji nada

   ‘You don't need it.’

   I:  no eta značit jizik ni [nada ili to [što ani [gavarili ni nada

   But does that mean [you] don't need the language or you don't need what they were talking about?’

   F1970s1: [da [da [da a eh vot (laughs) etava ja ni znaju što ana umela v vidu @ni nada i nji nada@ >my sibji dal’še igrali i my nji abraščali vnimanie a to što ani meždu saboj tam rasgavarivali<

   ‘Yes. Yes. Yes. Ah, there you go, (laughs) I don't know what she was thinking of. “You don't need it and you don't need it.” We went on playing amongst ourselves and didn't pay any attention to what they were talking about amongst themselves.’

In this case, it seems that the grandmother (whose Swedish-language conversation partners included her sister and son, both of whom were living in the same household) may have made a conscious decision not to speak Swedish to her grandchildren. However, two of her own children who survived the family's exile in the GULAG in 1945–1947 both grew up to be fluent speakers of Swedish. This may point to an interesting shift in the community's attitude towards language maintenance. Her last-speaker children were born in 1937 and 1943, during and after the Stalin-era repressions hit the village the hardest, and they are/were among the youngest in the village to grow up to be fluent speakers. Thus, this attitudinal shift may not have been connected to the active repressions, but rather to the events during and after the war. In any case, if there was such a shift, this interviewee seems to regard it as a response to higher-scale phenomena, such as interethnic relations and the explicit Sovietisation policies of the state (see e.g. Lane Reference Lane2010:72 on the connection between language shift and societal processes seemingly unrelated to language).

The notion of fear in the nontransmission of Swedish is, however, an interesting one. During my fieldwork in the village in 2013, I contacted most of the first generation non-Swedish speakers living in the village complex. Although many were happy to be interviewed, some did not want to participate. I discussed this with some other villagers, and several thought it was due to fear of repercussions “by the KGB”. It seems, however, that this too was an issue related to age; those who did not want to be interviewed were generally born around 1940. Again, this could mean that such fear was somehow linked to the events in the forties (cf. Wertheim Reference Wertheim2006 for a similar situation when doing fieldwork in another post-Soviet minority setting; the role of fear in language shift is well documented; see e.g. Campbell & Muntzel Reference Campbell, Muntzel and Dorian1989).

Yet, not all of the interviewees who talk specifically about nontransmission ascribe it to a conscious policy on the part of the adults. Another interviewee, an across-the-street neighbour and age peer of the interviewee in (2), thinks that it is simply lack of time and opportunity that is to blame for the nontransmission, as shown in (3).

1. (3) I:  što vy dumajete (.) pačemu švedy zdjes perestavali gavarit pa-švedski?

   ‘What do you think, why did the Swedes here stop speaking in Swedish?’

   F1970s2:  perestavali havarit’?

   ‘Stopped speaking?’

   I:  nu da (.) s::: djetmi

   ‘Well yes, with their children.’

   F1970s2:  (4.2) daže i nje znaju (1.6) m-mh.

   (pause) ‘I really don't know. (pause) M-mh.’

   I:  njet

   ‘No.’

   F1970s2:  daže i nje znaju ja

   ‘I really don't know, me.’

   F1970s2:   mnje kažetsja ščo sa rano v simje u kruhu svajom anji rashavarivali. prosta njekahda naverna byla učit djitej razhavarivat’. (1.8) naverna ja dumaju tak. može njehahda byla rashavarivat’ učit’ djitej. sso taki selo chazajstvo (.) rabota (.) večerom pridut ustavšie (.) može tak ni znaju. >a možna sami djeti ni chatjeli<.

   ‘It seems to me that within the family circle they did speak anyway. There probably just wasn't any time to teach the children to speak. (pause) Probably, that's what I think. Maybe there wasn't any time to speak and teach the children. After all, [in] the village [it's] farming, work, in the evenings they come home tired, maybe that's how, I don't know. Or maybe the children themselves didn't want to.’

In this answer, the interviewee talks about family life and how language transmission does not happen by itself if the parents do not use the language with the children. It is, however, notable that while she does not seem to cite conscious nontransmission as the reason for the children not learning Swedish, she also does not mention conscious transmission of Ukrainian. Rather it seems that in her analysis the parents did not speak much with the children at all, but that the children acquired Ukrainian outside the family. A similar view was expressed by a member of the parent generation, given in (4), who was born in 1937 and is thus one of the youngest to acquire competence in Swedish.

1. (4) F1937a:  å sås mama tå on ha:v ti:är så tåɽa hon då terheim me ås a fä:rr va:r mike arbet så ha:v mamma e- iŋa ti:är ent di tåɽa me ås

   ‘And so, mother, when she had time of course she spoke at home with us but in former days there was a lot of work so mother didn't have any time to talk with us.’

   I:  mm

   F1937a:  me lite tåɽa hon så en vi kont lä:ɖ ʂe lä:r ås sås

   ‘A little bit she talked so that we could learn learn that way.’

   I:  mhm (.) men mhh ha:v ni sistrar å bre:är hellär

   ‘Uhhuh, but did you have sisters and brothers or-’

   F1937a:  ja brö- brö:är ha:r ja

   ‘Yes, bro- I do have brothers’

   I:  mhm

   F1937a:  mhan bɽei de:är

   ‘But he died.’

   I:  aha men ni tå:ɽa milla jä:r på svensk

   ‘Okay, but between you you spoke Swedish.’

   F1937a:  ja.

   ‘Yes.’

This interpretation would mean that the language shift was not only a language shift, but part of a larger societal shift, where traditional patterns of socializing, working, and living were altered, resulting in language shift among other things. Considering the dramatic events the community went through towards the end of the 1930s and throughout the 1940s, this larger, social shift does not seem an unlikely explanation for the language shift. Furthermore, when the Swedish speakers returned to the village after the evacuation by force to Germany in 1943 only a few were able to settle in their original village, whereas many came to live in another, previously German, village in the municipality; thus the change in living patterns is clearly documented.

One interesting observation concerning the interviewees’ views on their parents not transmitting the language is that, whether the nontransmission was a conscious decision or not, most respondents talk about transmission in terms of actively teaching the children the language. It seems then, that in their perception, upholding a minority language requires a conscious effort; it is not necessarily that there was a conscious decision not to pass the language on (although some think there was), it is rather that there was not a conscious decision to pass it on (cf. Fogle & King Reference Fogle and King2013:2).

This is, of course, an important difference between the intergenerational transmission of a minority language and that of the default language of a society. Oftentimes, the default language (mostly the majority language) will be passed on implicitly, since society in general functions in that language and any member of it is expected to interact in it unless using another language is, at least, allowed. When it comes to a minority language, if the adults in the household are bilingual, transmitting the minority language requires either an active effort on their part or a community where interaction in the minority language is frequent enough for the children to acquire and maintain it outside of the family (cf. Fishman Reference Fishman1991:93–95, Caldas & Caron-Caldas Reference Caldas and Caron-Caldas2000). Thus, the aforementioned change in living patterns also meant that the already small Swedish-speaking community was further divided, in its turn making acquisition of the language outside of the home even more difficult.

The notion of language socialization outside of the home as a key factor in the Gammalsvenskby language shift is further substantiated by the fact that the new community language is Ukrainian. The vast majority of Ukrainian speakers in the area are evacuees or descendants of evacuees from Western Ukraine, a group which arrived in 1951 due to post-war border changes between the Soviet Union and Poland. This means that there was a clear increase in the proportional significance of the Ukrainian language on the local level around the time when the shift occurred. One interviewee pointed out that this community, too, has undergone a kind of language shift since that time. The local variety of Ukrainian is not the “clean Ukrainian” of the West, but Ukrainian with much Russian admixture, sometimes described by the villagers themselves as Suržyk. Where some of the last-generation speakers of Swedish or German spoke Russian as their main second language and the corresponding generation of evacuees speaks western Ukrainian, members of the next generation, whether ethnic Swedes or ethnic Ukrainians, generally speak this local variety of Ukrainian.Footnote ²

Day-to-day language use in a (non)speaker perspective

Turning to the second theme of this study, the issue of external pressure on the language-shifting group makes the reasons for studying language shift from the child generation's point of view even more obvious. In many of the interviews, first-generation non-Swedish speakers born between 1940 and the mid 1960s talk about derisive attitudes towards the ethnic Swedes and the Swedish language on the part of majority Ukrainians. An example of this can be seen in (5).

1. (5) I:  v vašej djetsve (.) ah drugie (.) djeti ani znali što vy n-nu švedka njemka?

   ‘In your childhood, eh, did the other children know that you are, well, Swedish, German?’

   F1960s2:  zna::li kanešna u škole vsje (.) nu::: prosto (.) jesli ščo nibud’ ni ta:k @ot švedi@ prativnaje abzyvali nas.

   ‘They did, of course, everybody at school, well, indeed if there was something not [just right] “There you go, [those] Swedes”, they insulted us with great loathing.’

And some last speakers correspondingly say that the pressure on the children to conform was one of the main reasons for the language shift, as in (6).

1. (6) F1930:  tå vi kåm från ti:skland (.) så broka dåm då sej åt ås faʃisti å ah- huʂʂ de vil (.) öikna:mas dåm åts

   ‘When we came from Germany, they used to say “fascists” to us and everything you can imagine. Call us names.’

   I:   kåfår

   ‘Why?’

   F1930:   tjå:: fere: feren vi va:r i tiskland så vild bå:na da:ʒe (.) terheim vild dåm ent en vi skild tåɽa (.) ent (.) eeh såɽe vild dåm ent tåɽa (.) mhm. en dåm- en eh (.) tåm a:r bå:na ent skild sej på tåm iŋatiŋ ent (.) ja (.) .ja:ja a såʂ dåm ha:vi lä:ɖ se

   ‘Ha! Because we were in Germany. And so the children didn't want us to speak even at home, they themselves didn't want to speak. Uh-huh! So that they, so eh the other children wouldn't say anything to them. Yes! Uh-huh. But otherwise they would have learnt [Swedish].’

At the end of (6), this mother claims that it was the children themselves who did not want to speak Swedish. As seen at the end of (3), this same notion also exists among the first-generation nonspeakers. In light of this, the interruption in intergenerational transmission would have been an intentional decision on the part of the children. As (6) also shows, some community members see this development as a direct response to ‘external’ pressure in the form of, for instance, name-calling (for a discussion on shame and language shift, see e.g. McCarty, Romero-Little, & Zepeda Reference McCarty, Eunice Romero-Little and Zepeda2006:668ff). Interestingly, however, this pressure, again, does not seem to have been directed at the language as such, but was perhaps rather ethnically motivated (see extract (5)). Such ethnic discrimination, in its turn, was based on the assumption that the ‘Swedes’ were enemies of the state, particularly since they had been evacuated to Germany when the German army retreated in 1943 (extract (6); cf. also Forsman Reference Forsman2015).

In much scholarship, the link between the heritage language and an ethnic identity is often seen as one of the most important reasons for language maintenance and revitalization (Fishman Reference Fishman1991; McCarty et al. Reference McCarty, Eunice Romero-Little and Zepeda2006). However, (5) and (6) suggest that the link between the heritage language and a certain ethnic identity—contrary to what Fishman (Reference Fishman1991:6) emphasises—may not automatically be an advantageous association, as, in addition to strengthening the potentially identity-constructing property of language for the minority-language speakers, it also opens up the possibility of discrimination on the basis of language as an index for other categories. In Gammalsvenskby, to use Fishman's words, to ‘be like everyone else, but in Xish’, might thus in fact have formed a better basis for the maintenance of Swedish (cf. Ó hIfearnáin Reference Ó hIfearnáin2015:59–60; for a more detailed discussion on language and identity in Gammalsvenskby, see also Forsman Reference Forsman, Kosunen, Lepistö and Rossi2014, 2015).

The oldest last speakers I have recorded (particularly those born at the beginning of the 1920s) told me that many members of the Ukrainian/Russian minority living in the village before the Second World War learnt to speak the local varieties of Swedish and/or German, and that there was no animosity between the different ethnic groups. Furthermore, while the accounts of derisive attitudes towards the Swedes are expressed as a matter of fact by those interviewees who were born in the 1950s and 1960s, those born in the 1970s claim—in an equally matter-of-fact manner—that they never experienced nor had ever heard of such behaviour. The difference in how the attitudes of the Ukrainian-ethnic population towards the minority are perceived by those born in the 1920s and 1970s on the one hand, and those born between the mid 1930s and the end of the 1960s on the other, may be linked both to the changes in the proportional sizes of the different ethnolinguistic groups, and to the events in 1937–1947. This seems to further support the conclusion that the language shift was, in fact, connected to a shattering of the sociolinguistic ecology, and that it was thus part of a much greater process of social change.

In view of this, it seems that the true question to be asked is one that is hardly ever asked in studies of language shift or, more naturally, in studies of families trying to create a favourable environment for bilingualism. Namely, is there an intrinsic value to language maintenance, and—more concretely, concerning the case of Gammalsvenskby—why should the Swedish community have maintained the heritage language? If, rather than evaluating the morality of the social circumstances surrounding the language shift in Gammalsvenskby or any other similar case, we want to learn about language shift as a social process, then the question is not whether external, social pressure was one of the reasons behind the language shift, but rather why the community responded to the social circumstances in the way it did. The case of Gammalsvenskby makes this question all the more central, since the outcome seems to have been virtually the same among the local German-ethnic population, and considering the observation by one interviewee, mentioned above, that even the community of evacuated western Ukrainians has undergone a language shift of sorts since settling in the area. In other words, the language shift among the Swedes in Gammalsvenskby does not seem to have been even a community-internal process, but a society-wide language shift from multilingualism towards societal monolingualism or, perhaps, Ukrainian-Russian bilingualism.Footnote ³

Returning to the issue of the value of language maintenance, many of the nonspeaker informants in Gammalsvenskby actually explicitly say that the value of the heritage language during the time period in question seemed to be close to none. In (7), this position is expressed by a second-generation nonspeaker, born in the 1960s, and thus an age peer to many first-generation nonspeakers.

1. (7) F1960s3:  tag- tagda: njebyla nikakova stsimula (pa)nimaješ mnohe: čitala @začem ana ono. začem ana ono@. (.) ne tuda: ni pajichat ni suda nichto ni priježav (.) a sečas ohho jakij stsi:mul ščo tse (.) priježajed i do naz hostji i i:: (.) ekskursij i (.) každe leta tuda ekskursij i ščaz velykij stsimul.

   ‘Back then there were no incentives, you see? Many thought “What [does] she [need] it for? For what [does] she [need] it”? No going there, nobody came here, but now o-oh what incentive, that it's- Guests come to visit us and, excursions and- Every summer [there are] excursions there and now [there is] a great incentive.’

This interviewee is among those who posit a direct link between the value of language maintenance and the extent to which the language provides a linguistic horizon beyond the local community. She points out that there was practically no contact between the Gammalsvenskby Swedes and any other Swedish speakers, and that the language thus did not bring with it any advantages that would have been unattainable otherwise. She also points to the situation after the breakdown of the Soviet Union and the fact that many Swedish speakers nowadays come to the village as tourists or to visit relatives. Today, knowing Swedish can generate income through small-scale tourism business as well as other employment prospects (e.g. in Sweden), and, indeed, the Swedish-language classes that have been available to both schoolchildren and adults during the past fifteen years or so, have been well attended not only by nonspeaking ethnic Swedes but equally by ethnic Ukrainians. The notion of Swedish as a practical tool for earning an income and gaining insights into the surrounding world is quoted as a central motivation by many of those nonspeakers who have taken part in Swedish-language classes.

Again, the same lack of linguistic horizon surfaces when some of the last speakers try to explain how the language shift came about, shown in (8).

1. (8) I:  å tåm lä:ɖ se ent (.) på svensk

   ‘And they [the children] didn't learn Swedish?’

   F1937b:  nej tåm lä:ɖ se ent på svensk. he: var iŋe enteresant ent. föʂtå: ɖe hu ɖe ska sej. nå nä- ti: å:r nudan he:r såm tåm vist en vi her le:va t-n- gamaɽsvenskar a såʂ kåɽe kåm te:r ehå:ne (.) en vi le:vd he:r. iŋen menski ent. iŋen menski kåm ent ihå:ne en vi he:r le:vd.

   ‘No, they didn't learn Swedish. It wasn't interesting. You see, how to explain. Well, when- [It's been] ten years now that they have known that we exist, the Gammalsvenskby Swedes, but otherwise, who remembered over there that we lived here? Not a single soul. Not a single soul remembered that we lived here.’

Once again, the mother in (8) seems to suggest that the language shift was actually something that the children brought about, and, as in (7), the reason quoted is that the language did not bring any added value. In fact, considering (5) and (6), the language would apparently have been a social disadvantage. Taking into consideration the political situation, where writing letters to friends and relatives had—at least previously—been construed as foreign espionage (Hedman & Åhlander Reference Hedman and Åhlander2003:261–62) and reading the Swedish books that were available, that is, mainly the bible and prayer books, was also tied to activities and a world view considered counter-revolutionary (Hedman & Åhlander Reference Hedman and Åhlander2003:279ff), many everyday language-maintaining measures were thus both socially and politically disadvantageous (on the relation between religion and linguistic identity in Gammalsvenskby, see Forsman Reference Forsman, Kosunen, Lepistö and Rossi2014, 2015).

This strand of inquiry, of course, shows again how individuals’ everyday linguistic practices are closely tied to their social circumstances. Admitting this does not mean that we accept that societies must inherently be unequal with some hegemonic discourses or cultures more worth maintaining than others. Instead, regardless of our personal ideological stances, we must accept that individuals act in response to their everyday reality, and the actions they take must thus be assessed within the framework of that reality. If we disregard the advantages of multilingualism that we as sociolinguists can think of, and if we look past the dismay we may feel when confronted with the atrocities of an oppressive political system, we must admit that there can be situations where the every-day reality of the individual offers no incentive for language maintenance. This may be all the more so in cases where the members of an old minority are bi- or multilingual and have been so for several generations; in such cases, the heritage language may not even be needed for vertical intra-group communication.

This, of course, means that ‘being like everyone else, but in Xish’ in Gammalsvenskby might actually have produced the exact situation that Fishman (Reference Fishman1991:6) associates with it, namely that ‘Xish really would be no more than a culturally redundant and cliquishly self-serving flourish’, which is precisely what the interviewees in (7) and (8) seem to say. However, if this is how participants in a language shift perceive the community language, then it is their right to do so—just as many would argue that it is their right to maintain the language, should they wish to do so (e.g. Hill Reference Hill2002:130; Skutnabb-Kangas Reference Skutnabb-Kangas, Skutnabb-Kangas, Phillipson, Mohanty and Panda2009:57). If the heritage language is not a pivotal index for the would-be speakers’ identity, as seems to be the case in Gammalsvenskby (cf. Forsman Reference Forsman, Kosunen, Lepistö and Rossi2014, 2015), and if the language does not bring with it any other added value (whether material or sociocultural), then no measures can make the potential speakers maintain it.