Engaging with Linguistic Diversity in Global Cities: Arguing for ‘Language Hierarchy Free’ Policy and Practice in Education

Abstract Educators working in diverse contexts and looking to promote language learning, as well as maintaining home languages, often encounter resistance and negativity directed towards a wide range of languages resulting in language loss at the individual level and language death at a societal level. On the other hand, a small number of languages are regarded as high status languages, in demand to study and speak. This paper has two aims: firstly, to identify processes which result in different attitudes to languages, which will be termed: language hierarchies and, secondly, to suggest an innovative way of creating conditions for development of ‘hierarchy-free education policy and practice’. In order to exemplify these processes the author’s previous research conducted in London and Cardiff is used. Hierarchy-free policy and practice is promoted as a necessary condition for sustainability of endangered languages. The second part of this paper introduces the concept of: ‘Healthy Linguistic Diet’ (Mehmedbegović 2011) and makes a proposal based on this concept for a new, cognitive-based approach to policy and practice in education which would help eradicate language hierarchies. This section presents latest research evidence from cognitive neuroscience, which supports the argument that all stakeholders in education and wider society would benefit from a shift in attitudes and approach to bilingualism and learning other languages. The benefits at the individual and societal level are so significant that acting on this evidence is not only an educational, but also a health and moral imperative too. In the conclusion, I give recommendations in terms of how this new strategy could be developed and implemented.


Introduction and key concepts
Being concerned primarily with regarding language sustainability, this paper is based on many years of engagement with London schools and bilingual pupils in different roles I performed: as a teacher, ethnic minority adviser, researcher and teacher trainer. Bilingual pupils in schools in England are considered to be all children exposed to (a) language/s other than English at home or in their communities. The Language Capital study (Eversley et al. 2010) maps 233 languages identified as home languages of children attending London mainstream schools. In addition, London as a global city is also a home to 14 international schools and a growing number of bilingual schools: French-English, Swedish-English, Italian-English, just to name a few.
Having had the opportunities to engage with policy and practice in all these different contexts I have been struck by a dichotomy that presents itself as a high-level awareness of the importance of learning and developing skills in several select foreign languages: French, Spanish, German and increasingly Mandarin, and a lack of recognition for the skills children already have in their home languages (languages spoken by parents, grandparents, often termed as heritage languages; however, this has always been for the purposes of being equipped for the globalised world and 'transnational careers' -as stated in the mission statements of many international schools that promote '"big' languages.). I have often found, when delivering professional development sessions in schools promoting maintenance and further development of skills in home languages, that educators argue: 'Our students only want to learn in English. Their parents want them to do well in exams. Their families don't want us to use home languages in teaching and learning.' Through my research on attitudes to bilingualism with parents, headteachers, lead professional and policy makers (Mehmedbegovic, 2008(Mehmedbegovic, , 2009(Mehmedbegovic, , 2011 I have been trying to identify first of all processes which result in this dichotomy of high status and low status languages, for which from now on I will use the term language hierarchies (see Ricento 2006, May 2012, Bilić Meštrić 2015, for various discussions of the idea) and secondly pedagogical approaches which would help create conditions for developing hierarchy-free policy and practice in education.
Hierarchies of languages for the purposes of this paper will be defined as ranking of languages in which certain small number of select languages are considered high status, desirable to learn and 'have', while a great number of languages are not seen as an asset and have a very low value status resulting in language loss at an individual level and language death at a societal level. This type of phenomenon is most commonly theorised through concepts of language and power. Language and power relationships have been analysed in many different historical and geographical contexts. Based on the review of relevant literature I would like to suggest the following classification of the four main processes which shape perceived and implied capital attributed to individual languages: process of capital accumulation, dynamics of the market demand, institutionalising discrimination of languages and political processes on the continuum from colonisation and oppression to political autonomy. I will proceed by discussing each of these processes.

Four processes 2.1 Process of capital accumulation
According to Bourdieu (2011: 81) "not everything is equally possible or impossible" in the process of accumulation that underpins the existence of all types of capital. Bourdieu writes about language as integrated in his notion of cultural and highly symbolic capital (Bourdieu et al. 2001). The recognition that the cultural capital may be converted into economic capital and 'institutionalised' in the form of education qualifications (Bourdieu et al. 2001) provides the basis for this layer of language hierarchies analysis. The inequality of languages in the context of convertibility into qualifications is immediately obvious in the fact that not all languages have access to pathways to gain recognised qualifications. For example, a GCSE in Albanian is not available in England, although there has been a growing number of Albanian speakers over the last twenty years. Shpresa, the Albanian community association in England, has been impressively proactive in organising high profile lobby events to raise awareness of the importance of having a GCSE in Albanian available. However, there has been no desired result achieved. In an article published by the TES (Times Education Supplement) titled 'New Fronts in the Battle for Minority Languages' a reference is made to the OCR exam board refusing the request to offer Albanian for the reasons of lacking a business case. The commercial criteria are critiqued by the academics involved and representatives of the Albanian community too, who object to the fact that skills children have in their home language cannot be officially recognised. In the same article Somali is mentioned as another language refused by the OCR exam board (Bearne 2011).
When we look at these two communities and their languages, both of them have been deeply steeped in political conflicts in their territories for a prolonged period (most Albanian speakers in England are from Kosovo). Both the Kosovan and Somali children arrive in the UK with the background of not having access to education in their countries of origin either at all or not in their home languages, for a variety of reasons: living in a conflict zone, not having right to open schools in Albanian, living in refugee camps in third countries. In the UK, a vicious circle of not being recognised by the system as a language with 'a business case' they are battling low values messages and the processes which clearly prevent them from every attempt to convert their cultural capital into education qualifications and ultimately their economic capital too. Although it seems that both communities are aware of the value of their languages and will continue to lobby for recognition, as a long term issue, the lack of recognition could lead to young people losing interest to maintain and further develop their skills in their home languages in order to focus on other areas which may be more useful in their future careers. As May (2012) puts it: groups that suffer from political marginalisation, social deprivation and economic instability are the ones that lose their languages.
Understanding the process of language capital accumulation based on the centuries of economic power, political domination and colonial gains, which have shaped the language capital of French, Spanish or English, in contrast to languages spoken by communities struggling for independence and international recognition, such as Kosovan Albanians, is necessary in order to appreciate the 'not-everything-is-possible' reality. Bourdieu (2011) writes about one's expectations and self-perceptions in terms of their place in a hierarchy of political power. The same processes can be identified in self-perceptions of young people in regards to language hierarchies. A 14 year old bilingual learner from a Bengali background attending a London school saying: 'Bengali has no value. Employers want French and Spanish.'(interview data, Mehmedbegovic, 2011) is a clear example of language capital self-assessment within the perceived language hierarchy, determined by market demands in this case. Market demands are discussed in the next section.

Dynamics of the market demands
Booming economies, developing business networks at the international level, working for global companies are some of the more obvious factors that shape the dynamics of the market demand for languages. However, business and economic growth are not the only areas of activity that require workforce with language skills. Humanitarian crises which are often less predictable, than future demands of the economy, place instant language requests on social services, health services, NGOs, foreign office, ministry of defence. One such example was the crisis in former Yugoslavia in the nineties, which resulted in thousands of victims of torture and ethnic cleansing being offered protection in many European countries, including England. Bosnian and Croatian languages became overnight highly sought after by services and agencies providing services to people rescued from the conflict zones and detention camps and to troops preparing to go to the conflict zone as a part of peacekeeping forces. Two languages which until then had very little value in terms of employability in England became languages in demand.
Military interventions and political conflicts similarly require immediate involvement of staff with relevant languages. September 11 th made languages such as Farsi and Dari, used in Afghanistan, of particular interest to military intelligence services. As more positive examples the bids for 2012 Olympic Games and the headquarters of the World Bank should be mentioned. Both of these used the language capital of London manifested in 233 languages spoken in London as an important argument and advantage of London in comparison to other applicant cities (Eversley et al. 2010).
These are a few examples that illustrate the point that the language market demands in different areas of work are influenced by events which range in terms of predictability from very high to very low. Market demands are the key process which will result in what Bourdieu (2011) identifies as convertibility of one's cultural capital into economic capital through increased employability. At the same time, market demands will also make their impact on the ranking of languages in terms of language hierarchy. Mandarin in England is an excellent example of a language which has risen in the last twenty years from the rank of a community language maintained within the community of Chinese origin families in community schools to a language that can now be seen on offer as a modern foreign language in some mainstream and many independent schools. Could it be termed: upward mobility: from a community language to a modern foreign language? In fact, this shift from a community language to a modern foreign language -means only something within the context of institutionalised discriminations of languages manifested in the division of languages imposed by the National Curriculum. Otherwise, Mandarin has always been a modern foreign language in the context of this country, but as a language that had not been offered on the National Curriculum and therefore not officially recognised, it was referred to as a community language taught in Chinese community schools to children of Chinese origin. Next, I explore further how the discrimination of languages is institutionalised.

Institutionalising discrimination of languages
The division of languages used in the education discourse in England, and therefore in London schools, separates languages into community languages and modern foreign languages is one of the main contributors to the process of discrimination and devaluing of many languages spoken by minority groups. This division of languages is historically based on the Orders for Modern Languages issued by DES in 1989. Apart from introducing foreign languages as a compulsory subject for the first time, it also classifies languages into two categories. The first category listed the languages of the European Community at the time, while the second listed languages spoken by large minority groups: Arabic, Bengali, Chinese, Gujarati, Hebrew, Punjabi, Russian, Turkish and Urdu. Schools were obliged to offer one of the languages from the first category, while all others were optional. Although the prescriptive nature of language offer has disappeared, it has been replaced by a greater freedom given to schools to make their own choices, many languages still suffer the burden of a lesser value label: community languages. My argument is that in order to depart from practices of institutionalised discrimination of languages it is necessary to identify the category 'community languages' as a label which devalues languages it refers to and supports hierarchical ranking of languages. This is manifested in the examples collected in a recent study: There are 76 languages spoken by children in my school and we operate English only rule on site. All these languages are invisible apart from Spanish and German as Modern Foreign Languages. The emphasis in schools should be on the importance of all languages. I was ridiculed for learning Swahili, now it is needed by the government. You never know which language is going to be important next. (Data collected with PGCE students, Mehmedbegovic et al. 2015) These two examples provide insights into experiences in London schools that on one hand marginalise or make invisible languages spoken by children and their communities while on the other hand devalue the attempts of English speakers to engage with community languages as a skill to learn or an academic subject. These examples are individual experiences and I am not suggesting that they are representative of London schools. However, my attempt is to present an alternative approach to all languages in education that is about creating conditions, discourses and understandings which have the potential to completely eradicate these types of experiences, and which are largely based on the cognitive approach. I will present this approach in the second part of this paper.

Political processes on the continuum from colonisation and oppression to political autonomy
For those interested in studying what can be done and undone to a language through means of political oppression, policies which can bring a language to the brink of its disappearance and the reversal processes of language revitalisation supported by political will and strategic approach -Wales is a case study par excellence. With the Act of Union between England and Wales, in 1563, Welsh lost the use and status of official language. English took over as the language of education, law and state functions, while Welsh was pushed out into the private sphere of family and community life. This oppression of Welsh remained unchallenged until 1942. However, it took nearly another twenty years of campaigning to get a Language Act (1967) which restored the official status of Welsh. The Language Act was followed by a period of another thirty years which it took for the full-scale language revival process to develop another Language Act (1993) (Ager 2003: 69). According to interview data with several lead professionals in education in Wales (Mehmedbegovic 2011), their parents chose not to teach them Welsh, since they could not see the value of it for their future education or employment. Now, being parents themselves and not speaking Welsh, they send their children to bilingual English Welsh schools. The value of Welsh has been restored, especially since the greater Welsh political autonomy in 1997. The language hierarchy where Welsh had no use in public services for several centuries has now taken a sharp turn and it has become a requirement for those working in the public sector to be proficient in Welsh too. The revival of Welsh has also been prominent in many other areas of work and culture, especially bringing up children bilingually from birth (‚Twf' 2006) and bilingual schooling.
A study conducted in Wales with 98 children age 8 -11 ) provides current insights into the attitudes to Welsh in a community of young learners in Welsh medium schools. The children who participated in this study reflect too on the fact that their parents do not speak Welsh and suggest that there should be games which would help their parents speak Welsh at home. A possibility of Welsh classes for parents was also mentioned in the focus discussion groups. These young learners also demonstrated awareness of how important bilingual English Welsh skills are in terms of their future jobs. They made suggestions for improving and increasing scope of activities in Welsh outside of school, but they were also very clear in requests to have more English as a part of their learning in school in order to develop academic proficiency in both languages. Most interesting and relevant to this paper is how even very young children in this study engage with the messages of language hierarchy and language capital as demonstrated in the following quote: We have been like born in Wales and everyone else you know, the people from France they speak French because they are happy of their language … Some people who come from Wales, they are just not bothered to like … not happy… (they) don't respect (the) Welsh country. (Thomas et al. 2014: 349) The case of Wales is an impressive example of a language being transformed from an endangered language, whose community of speakers suffered symbolic violence (Bourdieu 2001: 51) by internalising the devalued status of their language and contributing themselves to its disappearance, to the cultural capital that is now at the core of the co-constructing process of the contemporary Welsh identity. Although, it cannot be said that the lesser value associated with it in the past has fully disappeared, as the quote above indicates. In the words of one adult interviewee: 'it is a surprise that we Welsh speakers have survived at all, but now we are glad to be regarded as European rather than just British, it gives us more status as a nation and a language' (interview data, Mehmedbegovic 2011). The European Union as a wider community with a wide variety of languages, bigger and smaller communities of speakers, seems to provide a more reassuring wider context juxtaposed against Britain where the dominance of English can still be daunting especially to community groups who reflect on their languages by saying: 'we are surprised to have survived alongside English' (ibid. 2011).

Implications of language hierarchy on offer in education
In the climate of increased school autonomy, some schools have started to follow the trends of market demands with their language curriculum offer. French out -Chinese in headlines used in 2007 (BBC 8 O'clock news) referring to schools deciding to drop teaching French in order to introduce teaching Mandarin demonstrate the type of impact market demands make on education systems. However, seven years on, the British Council Language Trends Survey has identified that French, Spanish and German still dominate curriculum offer in mainstream schools (Board and Tinsley 2014). Primary mainstream schools in England mainly offer French and Spanish, while secondary schools widen their offer to include German, Arabic, Russian, Mandarin and Italian. Although, apart from German, all other languages listed are offered as an extra curriculum subject. In the independent sector, which includes international and bilingual schools (for a variety of languages alongside English: German, Italian, Swedish, French) curriculum offer includes a greater range of languages.
A language hierarchy manifests itself in the education system here as demonstrated by the findings of the British Council Language Trends Survey in two ways; firstly by an overwhelming dominance of French, Spanish and German as languages on offer and secondly by the status of these languages as fully recognised curriculum subjects, in contrast to all other languages being offered mainly as extra-curricular subjects (Board and Tinsley 2014).
A small range of languages on offer in the education system is only one side of the coin when it comes to the shortfall of the education system. According to the report on the state of the nation with regards to languages (Tinsley 2013), the deficiency of language skills is currently so severe amongst the available and new coming workforce that some large companies have started deleting language requirements from their adverts and staff profile requirements, having to focus their business strategies only on English speaking countries. State of the Nation Report terms this phenomenon "creation of a vicious circle of monolingualism" (Tinsley 2013).
I would like to argue that this is a process of perpetuating a well-established monoglot ideology in the domain on business and economy. In 2000, the Nuffield Foundation report (Nuffield Foundation 2000) was clearly highlighting 'the gulf' that existed between business language needs and education supply, which needed to be bridged for the benefit of the economy and individuals (Nuffield Foundation 2000: 18). This report was requesting a strategic approach to languages in education in order to overcome the 'vicious circle of monolingualism' clearly emerging from their findings too. The Nuffield Inquiry had a significant impact on policy-making arena. It initiated a parliamentary debate which led to appointing a National Director for Languages and developing the National Languages Strategy: Languages for All (DfES 2003). The thinking about how to best utilise existing language resources led to a further development in 2008 when the Institute of Education, London with a consortium of partners was commissioned to develop a National World Languages Strategy in order to widen the range of languages offered in education and utilise the languages skills of native speakers from minority ethnic groups.
Unfortunately, this strategic approach was stopped by then the newly appointed Coalition Government in 2011 and no further funding was invested into strategic efforts to utilise existing resources and secure at least baseline skills in languages for all learners. Conteh et al. (2007) write about language policies which serve to perpetuate long-established inequalities. In this case, the approach was no policy, no strategy from the Coalition Government. Status quo of well-established language hierarchy has been maintained by preventing the attempts approved by the Labour Government to develop a National Strategy that endeavoured to remove the division on foreign and community languages and replace it with engaged with the concept of World Languages with the aim of utilising existing linguistic resources represented by 233 languages spoken in London (Eversley et al. 2010).
Currently, in the existing policy vacuum the most important factor which results in an increased number of students studying languages in Key Stage 4 (14 -16), identified as a ten percent increase in a third of the schools in England (Board and Tinsley 2014), is the European Baccalaureate qualification, which includes a language. This is a clear message to learners that European qualifications and aspirations to become employable beyond UK borders require skills in languages other than English. Learners themselves are the focus of the next section.

Implications on learners
Current and recent national figures for England show that there is a difference of around four percent between the numbers of bilingual learners in primary and secondary phase, for example in 2012 there were 17.5 percent of bilinguals recorded in primary education in comparison to 12.9 percent in secondary education, while in 2009 a similar situation was recorded as there were 15.2 percent in primary education, while only 11.1 percent in secondary education (NCES, 2017). This means that around 40 000 children in mainstream education in England either lose their home language by the time they reach secondary school or deny being bilingual. A consultation conducted by the Institute of Education (2007) with experts in the field and regional advisers working with schools across England had identified that children from a variety of minority languages backgrounds often self-identify as monolingual due to deficit models attached to bilingualism in mainstream schools. The issue of opting for a monolingual profile and dismissing bilingualism as a resource is further supported by the evidence that many bilingual children often experience their home languages as of little value in the education system and perceive them of value only in their own communities (The Nuffield Foundation 2000, QCA 2006. My own research in London secondary schools (Mehmedbegovic 2008(Mehmedbegovic , 2009(Mehmedbegovic , 2011 provided insights into exactly the same patterns of schools being aware that a number of their students from bilingual backgrounds would insist that they were monolingual English speakers.
Looking at this phenomenon through the lens of the theory of misrecognition (Bourdieu 2001) this is a clear manifestation of 'symbolic violence'. Bilingual young people who have internalised the messages of marginalisation and devalued the status of their language capital are therefore opting for adopting what they see as the only affirmed language profile: monolingual English speaker as the product of the dominant monoglot ideologies (Silverstein 1996, Bloomaert et al. 2005, Bilić Meštrić and Šimičić 2017. As revealed in an interview with a bilingual professional currently working in education, who had a similar experience as a pupil, denial of bilingualism in her case led to the feelings of 'disinheritance' in terms of her own culture (IOE2007). In this case the process was supported by the decisions and actions of her parents who insisted on English only communication in their home environment too. Even though these parental decisions and actions were based on best intentions in terms of children's success in the education system of their host country, England, this participant was clear that denying bilingualism in childhood presents itself as a regrettable deficiency in adulthood.
The previously mentioned study with bilingual children in Wales (Thomas et al. 2013) provides the evidence of the pendulum swinging the other way and monoglot ideologies being used to prevent the use of the dominant language in favour of the minority language going through the revival process. The following quotes from a focus discussion group with young bilingual learners speak very clearly of the natural tendency of bilinguals to code -switch and how these processes are suppressed by the guardian of the language ideology in case -the teacher: We do Welsh but because he is … teacher not there, we don't have so much pressure so we use more English words. If we speak a lot in English someone will tell on us and we would be in trouble … so no one speaks English. (Thomas et al. 2013) Although it can be argued that the success of the revival programme and Welsh medium schools depend on facilitators of learning activities insisting on the use of Welsh, children feeling 'under pressure' and 'in trouble' for using their bilingual skills is essentially no different to their grandparents wearing a Welsh knot as a public shaming strategy when they used Welsh is schools in the 1940s and the 1950s. Working on re-establishing a marginalised language does not in any case justify the impact on individual learners, especially young learners, where they are experiencing fear and anxiety simply because they are drawing on all of the available resources they have in the process of learning. A monolingual dictatorship of any type: be it in support of a dominant or a minority language -only impedes learning and ultimately achievement and makes a negative emotional impact on learners. A bilingual head teacher I interviewed spoke of "a long term scar" he has carried for being told by his teachers that he could not spell because he spoke Polish at home (Mehmedbegovic 2008).
However, it is also important to point that monoglot ideologies are not only resulting in disinheriting and ultimately deskilling bilingual children and adults, they are also creating a false feeling of 'English is enough' spread amongst English monolingual speakers. This perception that English speakers do not need to speak other languages is based on the dominance of English unrivalled by any other language in our history: "no other language has been spoken by some many people in so many places … one in four of the human race is competent in English" (Crystal 2000: 10). In the parliamentary debate on languages which was triggered by the previously mentioned Nuffield Inquiry one speaker described this situation as 'drifting in the sea of English linguistic arrogance', while Lord Robson in the same debate identifies' language skills in other languages as a reason for failing to play full part in Europe' (Hansard records, 2002). The reality is that statistics on the representatives in EU bodies show substantial shortage of civil servants from the UK: a five percent of the European Civil Service are UK nationals, which is disproportionate to a 12 percent of the total EU population. The reason is that all employees need to offer either French or German as their second language and UK applicants do not satisfy that criteria (Tinsley 2013: 53).
A recently published Manifesto for Languages produced by All Party Parliamentary Group (June, 2014) argues that the UK needs to have a strategy for "national recovery in language learning"'. This Manifesto calls for a set of important commitments to be made by the political parties in the next general elections, which would form a Framework for National Recovery in Language Learning. In this document it is suggested that 48 billion could be saved to the UK economy if "national competence in languages is revived". This Manifesto exposes the of language learning decline at different levels, providing numerous examples: 44 UK universities stopped offering language degrees in the last 14 years, and that French and German A level exams have experienced a 10 percent drop in one year, to name a few.

Overcoming the hierarchy of languages
In the second part of this paper, I will outline my proposal for overcoming the previously analysed processes that result in a language hierarchy and its implications on education system and learners. My proposal is based on one unique aspect of developing bilingual skills which is of equal value regardless of the combination of languages. I would like to argue that this approach to languages and bilingualism could even overturn "not everything is equally possible" principle argued by Bourdieu (2011).

Equal values of all languages based on the cognitive aspect evidence
My main argument against "not everything is equally possible" in this paper is based on the findings that learning another language results in a considerable cognitive advantage which is evidenced in a range of studies covering all life stages from early years, compulsory education, adulthood and more recently even advanced age. A key point to this argument is that cognitive advantages of bilingualism are the same for any combination of languages: English-French; English-Bengali, English-Welsh, or indeed other options, which do not include English. In the continuation of this section I will present research evidence in support of his argument.
Clinical research studies carried out in the second part of the 20 th century involving bilinguals and monolinguals provide a significant body evidence which covers differences in a variety of variables, such as visual presentation and processing, audio processing, cortical activity of each hemisphere, levels of the right hemisphere engagement, levels of lateralisation and heterogeneity in the hemispheric organisation (Hamers and Blanc 2000). Jim Cummins in his book Power, language and pedagogy (2000) lists 160 education focus studies from different countries and contexts, all of which provide evidence that bilingual children perform better across the curriculum. Studies conducted with early years and school age children have found that bilingual preschool children demonstrate better focus on tasks while ignoring distractions than their monolingual peers. A similar enhanced ability to concentrate, a sign of a well-functioning working memory, has been found in bilingual adults, particularly those who became fluent in two languages at an early age. Managing two languages helps the brain sharpen and retain its ability to focus while ignoring irrelevant information (Bialystok 1999).
More recent studies with adults have provided insights into physical changes that happen on the brain when two languages are used demonstrated in the enhancement of the brain matter. Bilingual adults have denser grey matter (brain tissue packed with information processing nerve cells and fibres), especially in the brain's left hemisphere, where most language and communication skills are controlled. The effect is strongest in people who acquired a second language before the age of five and in those who are most proficient at their second language. This finding suggests that being bilingual from an early age significantly alters the brain's structure, but that the proficiency and intensity of use result in the same benefits (Kovelman et al. 2008). Similarly a study with of over 800 participants who were first tested as children in 1947 and retested as adults from 2008 to 10 found that: "bilinguals performed significantly better than predicted from their baseline cognitive abilities, with strongest effects on general intelligence and reading. The results suggest a positive effect of bilingualism on later-life cognition, including in those who acquired their second language in adulthood." (Bak et al. 2014a) This type of evidence is crucial in raising awareness on the importance of language learning as a lifelong activity while overcoming perceptions that learning a language beyond certain age might be too late in terms of impact on certain cognitive benefits. The 'Never too late?" study (Bak et al. 2014b) provides evidence that although there might be a difference in scope between early and late bilinguals in terms of enhanced attention switching, selective attention and auditory domain, both types show significant advantage in comparisons to monolinguals.
Bilinguals also show significantly more activity in the right brain hemisphere than monolingual speakers, particularly in a frontal area identified as the source of the bilingual advantages in attention and control. This expanded neural activity is so consistent on brain scans that it has been labeled as: "neurological signature" for bilingualism (Kovelman et al. 2008).
The latest evidence is even more significant in terms of one's well-being. Most recent research studies conducted in Canada point to bilingualism as a big hope in equipping ourselves better to deal with the threat of dementia: 'Executive brain power', developed by the use of two languages, has been identified as a key factor in prolonging quality life in later life and fighting off the onset of dementia by 3 to 5 years , Freedman et al. 2014.
Similarly, researchers from University of Edinburgh examined medical records of over 600 people in India and found that people who spoke two languages did not show any signs of three types of dementia for more than four years longer than those who were monolingual. They also established that bilingual advantage is not caused by any differences in education. Their illiterate participants with no formal education displayed the same benefits (Mortimer et al. 2014, Freedman et al. 2014. What implication do these findings have on our highest priorities for bilingualism in education? Furthermore, what evidence do we have that bilingualism enhances academic achievement? This is addressed next. Cummins (1976) has offered a theoretical explanation of the relationship between bilingualism and cognitive advantages that has to be fully proved by research, even though some evidence already exists. There is an interesting example, a study focusing on the relationship between bilingualism and mathematical abilities. Li, Nuttal and Zhao (1999) conducted this study with two groups of Chinese-American students, one group were literate in both languages and the other group not literate in Chinese. The group of students literate in Chinese achieved significantly better results on the mathematical tests for university entry.

Bilingualism and academic achievement
According to Cummins (1991: 84) the crucial elements that provide conditions for benefiting from the cognitive advantages of bilingualism specifically in terms of academic achievement are, first of all, that exposure to two languages provides broader linguistic experiences with the access to a wider range of thinking modes; secondly -switching between the two languages exercises flexibility in thinking, and thirdly, the conscious or subconscious comparison of two languages, resolving interference between languages, using the knowledge of one language to advance the other result in a high level of metalinguistic skills. This last point reflects the Vygotskyan view which says that bilingualism enables a child to see his/ her language as a particular system and to approach the language in a more abstract way and in more general categories (Vygotsky 1962: 110).
Metalinguistic studies that focus on the relationship between thought, word and meaning, again a key factor in Vygotsky's developmental theory (1962), shed light on the differences between monolingual and bilingual children. A number of linguists have used Piaget's sun-moon problem (1929) to test the ability of children to separate word from meaning and relate to word as arbitrary. This test consists in changing the names for sun and moon, getting children to decide which is up in the sky at night and finally what the sky is like at night: dark or light? In studies conducted by Cummins (1978), Bialystok (1988), Eviatar and Ibrahim (2000) bilingual and monolingual children alike accept with ease the name change and that sun would be what we see at night, bilingual children were quicker in making the final conclusion that the sky remains dark at night. Feldman and Shen (1971), Rosenblum and Pinker (1983), Ricciardelli (1992), Ben-Zeev (1997) conducted further studies getting children to use new names or nonsense names. All of these studies provided evidence that bilingual children either demonstrate higher flexibility in the use of newly agreed names or more abstract explanations for changing conventions.
The evidence of children approaching language and other academic content in a more abstract mode was recorded by two Canadian researchers, Lambert and Tucker, observing and testing a group of six-yearolds educated mainly in their second language. In this longitudinal study observed children demonstrated a high level of interest in comparing their two languages; approaching their second language as a code; using their first language as the basis for relating and translating both academic content and linguistic input. Therefore, the researchers were proposing that the acquisition of the second language had benefited not only the competence in their first language, but also their mastery of the academic content (Lambert and Tucker 1972: 82).
The research evidence referred to thus far leads one to question if neurophysiology can further add to our understanding of what the use of two languages means in the complex and invisible world of brain functioning. Uncovering the findings that neuropsychology has to contribute to having an insight in bilingualism certainly adds weight to the evidence provided by linguistic and sociolinguistic studies.
There are also studies which provide both types of evidence. For example, Ben-Zeev (1977) reports an identifying delay in reference to vocabulary and grammatical structures, while the advantages in terms of word manipulation, classification, structural analysis and nonverbal tasks in study comparing Spanish-English bilinguals and English monolinguals.
The most widely used theoretical framework which underpins identified good practice is the Interdependence Theory in conjunction with the Minimal Threshold of Linguistic Competence model of bilingualism as developed by Cummins (1976Cummins ( , 1978Cummins ( , 1991Cummins ( , 2000. The Interdependence Theory based on the concept of Common Underlying Proficiency (CUP) or the integrated source of thought for both languages accommodates the process of skills and knowledge transfer between the two languages, which is closely related to many positive findings on enhanced skills and cognition. Hamers and Blanc (2000) put an interesting perspective on this theory by defining the second language as "a function of competence in the mother tongue" (Hamers and Blanc 2000: 95). Concept of plurilingualism, as previously explored, employs Interdependence Theory and concept of CUP by promoting the principle of overall communicative skills which draw on all of our linguistic experiences.
On the other hand, the Minimal Threshold of Linguistic Competence conditions the positive transfer of skills between languages by suggesting that it is necessary to obtain a certain secure level of proficiency in both languages in order to be experiencing the benefits of bilingualism. On the contrary, delays and deficits in language development and use may occur. Bialystok, whose work is focused on language and cognitive development has contributed several studies supporting the view that the level of bilingualism determines its effect on development; higher the thresholds more positive the effects (Bialystok 1987(Bialystok , 1988(Bialystok , 1991(Bialystok , 2012. Cummins (2000) himself has reflected on the fact that both of his hypotheses have been misinterpreted and misused by policy makers, practitioners and supporters of both pro-bilingual education and antibilingual education. Cummins perceives his Threshold Model to be speculative and vague, because of huge variations that will depend on the environment, individual learners, languages, teaching methods. As the aspect most relevant for practice and policy-making in this area Cummins highlights 'the well supported finding that the continued development of bilingual children's two languages during schooling is associated with positive educational and linguistic consequences' (Cummins 2000: 175). This is also confirmed by Collier and Thomas (2007) study which provided evidence that the most reliable predictor of academic success in English is formal tuition in the students' mother tongue.
Cummins' message to policy makers leads into answering the second question: Is there a consensus amongst researchers regarding the advantages of bilingualism? The following quote from Bialystok captures what can be seen as the consensus in this area of research: bilingualism never confers a disadvantage on children who are otherwise equally matched to monolinguals and the benefits and potential benefits weigh in to make bilingualism a rare positive experience for children. (Bialystok 2012: 645) This concludes the section on the relevant research evidence. In the next section, I will argue for the policy changes which would eliminate deficit models of bilingualism in the education discourse and start a new chapter of utilising bilingualism for the benefits of individuals and society.

Education priorities
The research evidence presented in this paper underpins my argument that bilingualism is a source of cognitive advantage so significant at the individual and societal level that it is not only an education imperative to promote it, but a moral one too. According to the Alzheimer's society delaying the onset of dementia by five years would reduce deaths directly attributable to dementia by 30,000 a year. Dementia costs the UK £23 billion per year, yet the government has no plan on how to deal with dementia now or in the future (Alzheimer's Society, 2017).
If bilingualism is identified currently as the most promising strategy known in terms of prolonging a dementia free life, there certainly should be no delay in working on timely and necessary changes throughout the education system in order to benefit from its advantages.
For these changes to happen it is necessary to shift the thinking throughout the education system, from policy makers, to school leaders, practitioners, learners and parents. First imperative is that all the stakeholders are aware of what qualitative difference bilingualism can make to their cognitive functioning in general and specifically later life. Secondly, all stakeholders need to be equipped with strategies which can transform school practice and independent learning. This shift in policy and practice needs to start with school leaders and school workforce. By promoting development of school policies and practices which are crucial for an extended dementia-free life, we are at the same time offering enhanced cross curriculum performance for bilingual children and all those eager to learn other languages. Enhanced school performance is a key aspect in this shift which will be of interest to every school leader and teacher.

New strategy: Healthy Linguistic Diet
The concept of a 'Healthy Linguistic Diet' is based on the principle that all languages used by school children need to be supported in order to be maintained and developed further for the purposes of cognitive benefits. These benefits are not only needed by individuals from bilingual backgrounds, they are needed by schools too, as well as governments. For schools they mean better results in league tables, for governments they mean billions of savings in later life care.
The 'Healthy Linguistic Diet' I envisage as a strategy which would in the first instance: provide structured space for children and adults to discuss 'being bilingual' with the aim of raising awareness of benefits of bilingualism. The reason I start with explicit discussions about bilingualism is a 'critical incident' moment I had in a London school, which led to my conceptualisation of a Healthy Linguistic Diet. While facilitating a focus discussion group focusing on exploring reasons for underachievement with a group of Bangladeshi boys, one participant stated: "We underachieve because we speak two languages". His friend replied: "It is not true, I read in a scientific journal that bilingualism improves your brain." (Mehmedbegovic 2011). What struck me here as a critical incident was realising that 14 years old bilingual children were not given opportunities to learn about bilingualism in terms of research evidence and the impact of bilingualism on cognitive functioning. From this example it is visible that some children internalise the deficit model of bilingualism through a lack of any other model or explicit information on what it means to be bilingual, while those who are in search of knowledge about bilingualism have to do their own search of literature outside of expected interests and readings for young people in secondary education.
Based on this first-hand experience from an inner London school I have been working on developing principles and strategies which can be used for an approach conceptualised as thinking about a healthy diet -in this case it is a linguistic diet. Considering a big push for healthy lifestyles and healthy eating under the umbrella initiative Healthy Schools -I would like to suggest that concept of a 'Healthy Linguistic Diet' should be integrated into the National Healthy Schools Programme (Warwick et al. 2009). It has a real potential to contribute to its aims: raise achievement across the curriculum, improve long-term health, enhance wellbeing and improve inclusion. The following are suggestions for key whole school strategies: -Providing regular and rich opportunities for engagement and use of both or several languages: Teachers and school leaders need to find regular opportunities to provide consistent flow of affirmative messages with the aim of eliminating misconceptions about bilingualism as a problem and bad practices based on these misconceptions; -Providing access to and sharing relevant knowledge on values and advantages of bilingualism: Bilingual children and their parents need to be given clear, affirmative and consistent messages by schools and their teachers in terms of benefits of bilingualism and home language support. Students (and parents) should be given advice on what they can do themselves in order to support their own bilingual development. These messages should include raising awareness on cognitive advantages of bilingualism, which are applicable to all languages.
-Providing access to and sharing relevant knowledge on values and advantages of bilingualism: Teachers need to be provided with examples of good practice, guidance and training to develop skills essential for integrating home languages across the curriculum. This shift in practice should be led by the awareness that home languages used in teaching and learning will support bilingual children in developing their full potential. The impact of positive attitudes towards this specific intellectual potential that bilingual children have will be evident in improved results across the curriculum as a whole.
-Providing a framework which supports life-long development of bilingual competencies: All children (monolingual and bilingual/multilingual) and adults (monolingual and bilingual/multilingual) in schools should be encouraged to develop behaviour and habits which would support the life-long development of bilingual competencies. These practices need to become an integral part of our efforts to bring up children in the spirit of the Healthy Schools Programme leading to a healthy lifestyle.
The implementation of these key strategies is envisaged as driven by creative approaches such as bilingual theatre (as outlined in Kenner and Hickey 2008). They should be inclusive for bilinguals and monolinguals, children and adults, embedded in the school ethos and policy.

Conclusion
The National Audit Office estimates that nationally, dementia costs health and social care services £8.2 billion each year (Alzheimer's Research 2017). Alzheimer's Research UK has estimated that the overall cost of dementia to society as a whole is £23 billion a year. It has also been estimated that savings of £80 million could be made every year by improving hospital care. What savings can be made by promoting the 'Healthy Linguistic Diet'?
Loss of home language is often presented as a natural language shift or freedom of choice. I would like to conclude by emphasising that language loss is a loss for all who aim to achieve better attainment across the curriculum, better equipped workforce for a globalised world, better self-esteem at the individual level and longer dementia-free lives. Therefore, factors which shape language hierarchies and position certain languages as low value languages resulting in language loss and language death, as explored in the first part of this paper, need to be exposed and fought against like any other type of discrimination in our education system. The equality of cognitive benefits linked to bilingualism regardless of what combination of languages is in question needs to be promoted and explicitly communicated to all stakeholders: children, parents, carers, teachers and school leaders.