A history of sermons and carrots but no sticks

1 Introduction

This paper is concerned with educational policies on Modern languages or second foreign languages (SFLs) like French, German and Spanish in Sweden. Foreign language education in Sweden has a strong tradition. As a country with a comparatively small national language, the need for Swedes to master foreign languages has been acknowledged for a long time (Cabau-Lampa 2005). Consequently, there is a need for a well-developed and active language policy. Indeed, in some areas Sweden has a long-standing tradition of an active language policy (Oakes 2001; Hult 2004). The Home Language Reform, orginally implemented already in 1977/1978, states that immigrant children have the right to study their mother tongue in Swedish schools (Boyd and Huss 2001). Sweden signed and ratified the European Charter for Regional or Minority Languages in Year 2000. Together with a heated debate on threats to Swedish as the national language (e. g. domain losses to English, competence attrition, signs of diglossia, Teleman and Westman 1997), the signature lead to the Language Act (Ministry of culture 2009) which came into force in 2009. It states, for the first time ever, that “Swedish is the principal language in Sweden” (section 4) and that “The national minority languages are Finnish, Yiddish, Meänkieli (Tornedal Finnish), Romany Chib and Sami” (section 7).

Foreign languages are not mentioned in the current Language Act. In the Government bill adopted in 2005 – entitled Best Language: A Concerted Language Policy for Sweden (Bästa språket: En samlad svensk språkpolitik) – four objectives for language policy in Sweden were legislated:

Everyone is to have the right to language: to develop and learn Swedish, to develop and use their own mother tongue and national minority language, and to have the opportunity to learn foreign languages. (Government bill 2005/06:2: 14)

However, in the Government bill that a few years later became the Language act, it was concluded that there were “not enough reasons” (Government bill 2008/09:153:43) to include the right to learn foreign languages in the Language Act. Instead, foreign language policy is and has always been a concern for the educational system in Sweden, as is the case in most countries according to Spolsky and Lambert (2006). Compared to language policies in other areas, Spolsky and Lambert find that foreign language policies are:

[...] less developed and conflicted than, policies with respect to national language(s) and those of intra-country minorities. They also tend to be given less attention in scholarly analyses of language policy. In addition, such policies tend to be piecemeal rather than coordinated. (Spolsky & Lambert, 2006: 570)

Indeed, when considering the vast literature on language-in-education policies (e. g. Tollefson, 2002, Liddicoat, 2013) or language acquisition management (LAM) (e. g. Baldauf et al., 2008) there is comparatively little research focusing specifically on foreign language policies in education (but see e. g. van Els, 1994, 2005, Mitchell, 2009 for notable exceptions). Instead, current theoretical discussions of LAM typically do not separate between ‘types’ of languages (i. e. national, minority, heritage, foreign etc) even though some policy areas are potentially more relevant to foreign languages than others. One such area identified by Kaplan & Baldauf (2005) which will be central to this paper is access policy, i. e. the question of who should study which language when and for how long. With respect to second or foreign languages, van Els (2005) argues that access policy should be based on a needs analysis at the national level:

The second languages that are provided for in the educational system should be selectd in such a way that the country’s current and future needs for second languages are satisfied in optimal ways. (van Els, 2005: 974)

In her review of contemporary foreign language education policies, Mitchell (2009:89) does indeed find that at least in the European context, it is clearly instrumental goals that are currently driving the development of foreign language education policy with respect to language choice(s) and amount of time devoted to foreign language study. Van Els takes as his point of departure the traditional nation-level second/foreign language status planning (van Els, 2005: 972). For reasons of space, the mangagement perspective will also dominate in the current paper even though current models of language policy are more complex and acknowledge agents at different interacting levels. Spolsky’s (2004) model of language policy involve three inter-related components: practices, beliefs and management. Practices are concerned with behavior and choices that language users make and beliefs about language(s) are important for the values that we assign to language(s) and ultimately for the construction and maintenance of identity. Spolsky (2007) shows that within a particular domain (e. g. home, school, church etc) each of these three components will exerce pressure on language choices. With respect to the school domain, Spolsky (2007: 7) says that it “proves to be the most complex” with different participants (i. e. students, teachers, school leaders etc) and authorities at different levels (i. e. national, regional, municipality, school) exerting influence on choices and implementation. However, Spolsky (2007: 11) finds that, in the end, “schools reflect the ideological position of those who control them” and it is rare that linguistic or educational considerations will ultimately determine the language policy. As a result, conflicts in the school domain are frequent. The emphasis on conflict or rather its opposite – consensus – is echoed in the conclusion of Mitchell’s (2009:98) review of contemporary foreign language policies in education when she finds that successful foreign language policies ”cannot be imposed by ‘bureaucrats’ and ideologists, but require a reasonable consensus among students, parents and the wider community to take root and deliver a measure of foreign language learning” with English as a foreign language as the current most prominent example of a success story.

The current paper is concerned with an historical review of acess policies with respect to SFLs in the Swedish educational system. I will mainly adopt a management perspective at the national level, but with some outlooks to practices and beliefs of agents at other societal levels as well. While policy decisions and objectives will be central to the review, I will also be focusing on the instruments that Governments have used in order to achieve these goals. Since SFLs have been and still are a concern for education, the policy instruments available to the Government in order to regulate access policies are educational policy instruments. Policy instruments or policy tools can be defined as combinations of “practical, intentional and administrative actions” (Salazar-Morales 2018) that governments employ to achieve policy goals in specific social domains. Bemelmans-Videc, Rist & Vedung (1998) have proposed a widely cited threefold typology to describe the different instruments that Governments normally have at their disposal: persuasion, incentives and controls or, using language that is more colloquial, sermons, carrots and sticks. In brief, persuasions or sermons involve informing the target population about facts related to the issue at stake through communication campaigns, labelling or educational efforts to make people understand the benefits of a particular behaviour. It also involves informing and reminding the public about what the policy target is, i. e. the objective to achieve. Differently from persuasion, Governments use incentives or carrots to induce change in a target population through the conditional transfer of some kind of value, often of economic nature (e. g. exemption of taxes). A fundamental assumption behind the use of incentives is the idea that some people are “utility maxi-misers” (Salazar-Morales 2018) and are not likely to change behaviour solely on the bases of exhortations. Finally, controls or sticks is the most constraining public policy tool in the toolbox and uses authority-based coercion as its basis. In the domain of educational policy, all kinds of compulsory measures can be qualified as sticks. Crucial to the success of sticks is the perception of legitimacy and support among the public (Salazar-Morales 2018: 461). If people do not accept the restrictions of the authorities, they stand little chances of being accepted or maintained over time.

The aim of the present paper is to review access policy goals with respect to SFLs and the instruments used to obtain them after 1945 in Sweden and to discuss them within the framework of policy instruments of Bemelmans-Videc et al (1998). The focus will be on compulsory school, which in the case of SFLs means lower secondary school (age 12 to 15 years). Looking at the period from 1945, the research questions are threefold:

1. What were the objectives of the educational policy with respect to SFLs?

2. Which type of policy instruments were used and why?

3. What were the consequences of these instruments?

The history of foreign language education and foreign language teaching in Sweden has been the topic of several previous publications (Tornberg 1997; Cabau-Lampa 2005, 2007; Tholin and Lindqvist 2009; Tholin 2019). Therefore, I will only quickly review the earlier periods focusing on facts needed to answer the research questions (Section 2). Comparatively less has been written on the situation after year 2000 when the Swedish Government introduced a new type of measure. Consequently, I will discuss in more detail the current situation and consequences of the current policy (Section 3). The discussion and conclusion (Section 4) will summarize the findings and discuss to what extent the underlying ideology with respect to SFLs in Sweden has remained the same over the years.

2 A brief historical overview of second foreign language education in Sweden

Second foreign language education in post-secondary war Sweden can roughly be divided into three periods. The periods are defined following the introduction of new curricula involving major changes with respect to SFLs. The focus of this overview is the educational system, policy goals and the instruments used to achieve them. For reasons of space, teaching and assessment methods will not be included (but see Cabau-Lampa 2005 and 2007 for details).

3 The consequences of the GPAEC system and the current situation for SFLs

A crucial question is to understand if the GPAEC incentive measure works. Does it have the intended effect of raising the proportion of pupils who study a SFL? To date there has not been any comprehensive evaluation of the GPAEC measure, but the effects of this measure have been commented on in different reports and papers.

The European Commission (2014) published a country-comparative report as a follow-up to the First European Study on Language Competencies study. For Sweden, the report finds that the introduction of GPAEC has increased the completion rate and the number of pupils studying SFLs in the country, but there is no data in the report to support the claim. The National Agency for Education (2018) discusses the effects of the GPAEC in upper secondary school and finds that more pupils complete higher levels of SFLs after the introduction of the GPAEC system than before. Since it is not possible to complete higher levels of SFLs in upper secondary school if you have not started studying the SFL already in lower secondary school, the National Agency for Education finds that the introduction of the GPAECs for SFLs should also have had a positive effect on the proportion of pupils studying a SFL in lower secondary school.

A more thorough analysis of the trends in the study of SFLs before and after the introduction of the GPAEC system was carried out in (Granfeldt et al. 2020). This study looked at the proportion of pupils studying a SFL in the last year of lower secondary school (Year 9) between 2000 and 2018 and compared the period before and after the introduction of GPAECs. The results showed that on a national level around 64 % of the pupils studied a SFL in Year 9 in 2007, the year before the introduction of the GPAEC. This figure had raised to 72 % in 2018, an all-time high for the 21st century. It would seem, then, that the GPAEC had a positive effect in the way it was intended. However, Granfeldt et al. (2020) show that the increase is not equally distributed across different demographic municipalities. In fact, in Year 2000, the proportion of pupils studying a SFL in Year 9 did not differ in large cities, medium-sized towns and smaller towns/rural municipalities. In all three demographic groups around 62 % studied a SFL. In 2018, 76 % of the pupils in large cities studied a SFL and 66 % in medium-sized towns. In smaller towns and rural municipalities, there was no increase at all between 2000 and 2018. There are two important observations in this data. First, it means that the increase at the national level after the introduction of the GPAEC system is mainly an increase in large cities and to some extent in medium-sized towns. Second, it means that what was once a relatively homogenous situation in the sense that there was no or very little difference between urban and rural areas with respect to the study of SFLs has no become much more differentiated.

The question is what might have caused this differentiation. While the answer is not straightforward with the type of data we have at hand, it would seem plausible that the GPAECs has played a role in this development. This becomes even more apparent when statistics concerning the transition to post-secondary education are considered. Recall that the GPAECs for higher levels of SFL courses increase the possibilities of entering university. The proportion of young people who continue studying at university is generally higher in large cities and municipalities near large cities than elsewhere (Swedish Higher Education Authority 2016). In other words, it is in precisely those demographic regions where the expectancy to go to university is the highest, that the study of SFLs in Year 9 has increased the most.

However, there are reasons to believe that the demographics per se is not the most important part of the puzzle. Krigh (2019) investigates the study of SFLs from an educational sociology perspective. Working in a Bourdieuian framework and drawing on the notion of capital, Krigh asked the question of which pupils with respect to gender, social and national background and grade point average that chose to invest in the study of modern languages (Krigh 2019: 221). Most pupils start studying a SFL in Year 6 or Year 7, but boys more than girls abandon the subject before the end of lower secondary school (cf. also Tholin and Lindqvist 2009). Native-born pupils more than foreign-born invest in a SFL. Pupils coming from lower middle class and working class abandon their SFL most. Among working class boys, only one in two studied a SFL in Year 9 in Krigh’s data. The gender difference was also the largest among the working class families. Pupils coming from families with a high cultural capital were the ones who received the highest grades. In conclusion, Krigh finds that pupils coming from resource strong families are the ones who are the most likely to persist with their studies of a SFL until the end of lower secondary school. Krigh has also noted the all-time high proportions of pupils who currently study a SFL in lower secondary school and points to an other interpretation of the “language crisis” for SFLs:

Therefore, there are reasons to conclude that the language crisis, as presented in public debate, does not exist. On the contrary, more pupils than ever study modern languages in compulsory school. On the other hand, there is reason to discuss a potential inequality crisis in relation to which groups of pupils participate in the elective educational pathway that studies in modern language offer within the framework of a cohesive compulsory school – an elective pathway which disadvantages the daughters and in particular the sons of the working class because it favours upper and middle class children, in particular their daughters. (Krigh 2019: 240, my emphasis)

The GPAEC educational policy instrument seems to have fitted very well with the groups of pupils who traditionally choose to study a SFL in school. The carrots introduced by the Government probably corresponded to the needs, the interests and the intended educational investments of the upper and middle-class children and their families. The consequences are all-time high levels of participation in SFL classes in lower secondary school, but at the same time an increasing social differentiation between those of who study more than one foreign language in Sweden. The inequalities have grown and the EU idea of multilingualism for all (beyond English) is increasingly more distant.

Another clear trend of inequalities in the Swedish SFL landscape during the last 20 years concerns the diversity of the SFL offer across the country. Again, it can be argued to be a consequence of the educational policies in the domain. Sweden is among the countries that have experienced a ‘Spanish boom’, i. e. a considerable increase in the interest in studying Spanish. Since Spanish was introduced as an option in lower secondary school in 1994, the language has seen a dramatic increase in the proportion of pupils choosing it. The increase took off in the beginning of 2000 and has led to a situation where, today, more pupils choose Spanish than French and German together (Francia and Riis 2013, 2016; Granfeldt et al. 2020). Interestingly, Spanish was the most frequently studied SFL in lower secondary school in all social groups according to Krigh (2019: 224). While Spanish has seen a sharp increase, German saw an initial sharp decrease in the beginning of the 2000 s and lost its position as the most studied SFL in 2006. After 2008, the proportion of pupils choosing German has been rather stable just below 20 %. French saw a smaller decrease in the beginning of century, but has stayed relatively stable around 15–17 %. Importantly, the SFLs are not chosen to the same extent across the country. Granfeldt et al. (2020) found that while Spanish is the most studied SFL everywhere, French is mostly studied in large cities. German on the other hand is comparatively more studied in small municipalities and rural areas. In small municipalities with few students, French has traditionally had its weakest support.

In the current School ordinance (Ministry of Education 2011), organisers are required to offer two out three of the languages French, German and Spanish and they should strive to offer more. In other words there is a stick saying that an offer of two SFLs is a minimum and sermons saying that organisers should offer three SFLs or even more. A clear majority of schools still offer all three languages (Granfeldt et al. 2019), but there is an increasing number of municipalities that do not register any pupils in one of the languages. This is particularly the case with French, which has seen a quite dramatic increase in municipalities where no pupils study the language, in many cases because of a reduction of the SFL offer from three to two languages (Granfeldt et al. 2020). Table 1 below shows that in 2007, the year before the introduction of GPAEC, there were 9 municipalities (out of 290) or 3 % without registered pupils in French in lower secondary school. 12 years later the same figure is 46 or 16 %. Municipalities representing smaller towns in rural areas are overrepresented in this category. Table 1 also shows the continuous spread of Spanish. In 2007, 40 municipalities had no pupils in Spanish, a figure which is down to 4 in 2019. German has only seen a small increase in these statistics, which might be because comparatively more pupils choose this SFL in smaller towns and rural areas.

A current trend is that some organisers reduce the offer of SFLs to Spanish and German. This means that it is increasingly often the case that the offer of SFLs that pupils have depends on where you live in Sweden. If you grow up in a small rural municipality, there is a high risk that you cannot study French in school. Stakeholders and the public have recently reacted to this progressive trend of inequalities in the media (see Krigh 2019: 11 for a discussion). In other words, the organisers at the municipality level are complying with the stick, but less and less with the sermons coming from the Government.

4 Discussion and conclusion

This paper is concerned with educational policies regarding SFLs in Sweden with a particular focus on access policy, i. e. the question of who should study a SFL when and for how long. The historical overview shows that both policy objectives and instruments used to reach the objectives have changed since the 1940 s. This is not surprising. Societies have gone through dramatic changes in all social domains during this period and Sweden is obviously no exception. It is perhaps more surprising to find that there are resemblances between educational policies for SFLs from 80 years ago and the ones currently in vigour. The relationship between English and SFLs like French, German and Spanish has remained the same over the whole period. On the one hand, English has from the outset been governed with (increasing) control measures. Successively, English has been introduced at an even earlier age and it has been mandatory for all to study the language, despite initial doubts among an important group of agents, the teachers, about the feasibility of ‘English for all’. On the other hand, the elective status of SFLs like French, German and later Spanish has been fundamental in educational policies over the whole period, despite political and expert voices raised since the 1960 s to make the subject mandatory in compulsory school. The type of policy instruments with respect to SFLs has varied, but the target population has essentially remained the same. In the earliest period, SFLs were officially regarded as a screening instrument, a way of identifying gifted pupils who could continue their education at higher levels. In the beginning of the 1960 s, SFLs were part of the gateway to upper secondary school. When this role is lost in the democratization process of the late 1960 s, the result is a confusion regarding the status of SFLs and a dichotomous debate between extreme positions – exclude or make mandatory – emerges. With the entrance into EU the democratization argument – multilingualism for all – is reinforced through the ‘mother tongue + 2’ language policy, but the Government is not prepared to choose any sticks at this point, possibly because neither the SFL teachers nor the public are convinced. Instead, the Government chooses persuasion or sermons to stress the importance of SFLs as Sweden becomes a member of EU.

By choosing persuasion and not control as their policy instrument, the Government should be aware and should accept that not all pupils will study a SFL, but the Government still becomes concerned when there is a perception among stakeholders that too few pupils actually do so. This was the case in the beginning of the 2000 s and the reason why the Government chose to introduce yet another type of policy instrument: a carrot in the shape of GPAECs for SFLs. The idea behind incentive measures is that people are “utility maxi-misers” (Salazar-Morales 2018). They will analyse the incentives and see how they best can profit from them. The GPAEC system first targeted the transition from upper secondary school to university and then the transition between lower and upper secondary school. Available analyses seem to indicate that there was a good match between this measure and social groups where there was already an intention to invest in further education. The results have been all-time high levels of pupils studying a SFL at the end of compulsory school, but also an increasing social differentiation between pupils from upper and middle-class who possibly more than ever study a SFL and those from working class families who do not intend to continue to university to the same extent.

The question is how Sweden ended up with these national ping-pong policies with respect to SFLs and why sermons and carrots, but much less sticks have been favoured over the years. In the remaining section of this paper, I will discuss two interrelated factors: the status and the proficiency in English and the lack of a foreign language policy in Sweden for languages other than English.

The early decisions to make English the first foreign language and mandatory were important. Cabau-Lampa (2005, 2007) and Tholin (2019) both argue that English was regarded differently from SFLs from the start. The argument for English was based on a needed utilitarianism, because of the internationalization efforts after WW2. English was never meant to play the role of screening instrument which was the case for SFLs and in particular German. The National Board of Education was less worried that some pupils would not be suitable for learning English, despite what the teachers felt (Tholin and Lindqvist 2009). Moreover, there was probably already in the 1940 s a perception among broader layers of society that English was useful (Johansson, 2004). A second key moment is in 1994 when there was again the possibility of making SFLs obligatory, but by that time the view of ‘English is enough’ is probably well instituted in the society and there was not enough support in the public to introduce any control measures for SFLs (Cabau-Lampa, 2005). Since then, several reports seem to have supported the view that language proficiency in English is generally high (e. g. European Commission 2011) and the language enjoys a very strong support among the public (European Commission, 2012). In conclusion, the early and strong educational focus on English as the first foreign language together with the strong support and high proficiency in the language have made any sticks impossible with respect to other foreign languages. Salazar-Morales (2018) points out that control measures need public legitimacy in order to succeed. This has never been the case for SFLs.

The second factor is the lack of a clear foreign language policy in Sweden with respect to other foreign languages than English. There has never been a needs analysis with respect to foreign languages or foreign language competences in Sweden as in other countries (e. g. the Netherlands, van Els 1994, 2005) even though several stakeholders and experts have pointed out this lack (e. g. Association of Swedish Higher Education Institutions 2015). This means that there is no available knowledge base on which an educational policy could be constructed. In Sweden, there is no research-based knowledge about the economic, educational or cultural benefits of investing in SFLs for different sectors of society or for the individual even though some published reports indicate that Swedish compeitiveness is suffering due to a lack of foreign language proficiency in small and medium-sized companies (Rosén, 2017). There is a need for more research in this domain.

Another consequence of a lacking foreign language policy is that it is very difficult to evaluate whether a current policy and the accompanying instruments are successful or not since the goals are not specified. This became apparent in the beginning of the 2000 s when the number of pupils studying and completing a SFL decreased. Even though the elective status of the subject and the persuasion tool that accompanied it would predict that all pupils will not choose a SFL, the Government became worried and they chose to introduce the incentive measure of GPAECs, possibly as a reaction to the emerging debate on a ‘language crisis’ at the time. What a satisfactory proportion of pupils studying a SFL in the country is has never been clearly layed out. Currently, there is an all-time high proportion of pupils studying a SFL at the end of lower secondary school, but the discourse on a ‘language crisis’ continues. However, it is highly probable that the GPAEC carrots have been effective, but they were mainly utilised by the same population that has always been the target of Swedish SFL education policies, pupils from upper and middle-class families.

In conclusion and looking at the period from 1946 of educational policies and instruments in the domain of SFLs, there is in fact a certain consistency. Sermons have not been perceived to be enough, carrots tend to lead to inequalities and sticks have always been avoided.