Albanian-Greek border crossing at Kakavijë (Source: greece.com)

Our understanding of the role of language in social life suffers from a particularly intractable problem: the terms we use to speak about language are often not very useful; on the contrary, they are confusing, obscurantist and create problems that we cannot even see because of our limited linguistic imagination.

Schools in particular frequently fall prey to applying imaginary language labels to their students. Last year, we reported on an Australian study that found that only two out of 86 schools in Queensland felt they held accurate data about the language backgrounds of their students. This shocking finding was due to the fact that language assessment was conducted in haphazard ways; and assessors were prone to confuse ethnic identity with language proficiency. It was also due to the fact that they had no useful language labels at their disposal, with “English” and “language background other than English” as their main categories. This meant, for example, that creole speakers would be labelled as native “English” speakers while monolingual Standard English speaking children might be labelled “language background other than English” just because they had a migrant parent.

The problem with such messy records is not only an academic problem with data accuracy; much more importantly, the consequences of these dodgy data are that the role of language proficiency in learning is consistently misrecognized: creole speakers who are labelled as native English speakers may be considered slow learners and, because their language problems are not recognized, they do not receive the language support they need.

A recent Albanian study provides a further example of the ubiquity of linguistic misdiagnosis and its detrimental consequences. The researchers, Zana Vathi, Veronika Duci and Elona Dhembo examined the schooling experiences of “returnee” children in Albanian schools.

Albanians in Europe (Source: wikipedia)

Albania is one of the poorest countries in Europe and a country with a very high emigration rate. The preferred international destination of Albanian emigrants is neighbouring Greece. However, as is well known, Greece was hit particularly hard by the 2008 global financial crisis. Within Greece, Albanian migrants were particularly vulnerable and around a quarter of adult Albanians returned from Greece to Albania in the period between 2009 and 2013 (134,544 out of around 600,000 Albanians in Greece).

Many of these returnees had school-aged children who needed to transition from a school in Greece to one in Albania. Most of these children were Greek-dominant bilinguals with varying levels of proficiency in Albanian. However, the fact that they are ethnically Albanian means that they are expected to speak Albanian natively, in the same way that their monolingual peers who have only ever been schooled in Albania do. In other words, ethnicity obscures language proficiency.

As a result of this mismatch, the educational experiences of returnee students were mostly negative. Many reported that their language was often the object of ridicule, not only by their peers but even by their teachers. With the exception of a school in the border region which also catered to a local Greek minority, none of the schools in the study were prepared to cater for the specific linguistic needs of returnee students. Measures such as supplementary Albanian classes for this cohort were non-existent.

While schools failed to recognize the specific linguistic needs of Greek-dominant bilingual students, they imagined (and addressed) another “need” they thought returnee students had: they imagined that the returnee students were particularly in need of a patriotic education. The consequence of such well-intentioned efforts to support their “Albanianness” constituted a further source of exclusion: many of the students missed Greece, considered themselves to have plural identities and had global rather than narrowly national aspirations. While teachers thought they were being supportive by strengthening the students’ Albanian identities, returnee students felt deeply alienated when only one aspect of their multiple identities was valued.

There is a widespread assumption that the central problem in the education of migrant students is related to ethnicity. As this study clearly shows being part of the dominant ethnicity does not protect migrant students from the effects of the linguistic and institutional disjuncture between different national school systems. As a result of this disjuncture, most returnee students experience downward educational mobility: they move into lower grades than those they had attended in Greece and their performance worsens. Some drop out of school altogether.

As long as we do not even have terms that could disentangle linguistic proficiency from ethnicity, schools around the world will continue to fail students who do not quite fit the profile we imagine for them.

If you would like to read more about the ways in which linguistic diversity shapes the educational experiences of minority students around the world, you might be interested in Chapter 5 of my new book Linguistic Diversity and Social Justice, just out from Oxford University Press

Piller, I. (2016). Linguistic Diversity and Social Justice. Oxford University Press. DOI: 10.1093/acprof:oso/9780199937240.003.0005
Vathi, Z., Duci, V., & Dhembo, E. (2016). Homeland (dis)integrations: Educational Experience, Children and Return Migration to Albania International Migration DOI: 10.1111/imig.12230

Barely legible today but evidence of ‘super-diversity’ in the 9th century: Runic graffiti in Hagia Sophia

On a parapet in Hagia Sophia’s gallery there is an obscure little graffiti written in Viking runes and dating back to the 9^th century. All that is legible today is ‘alftan,’ which refers to the Norse name ‘Halfdan’ and it is assumed that it was part of a formula such as ‘Halfdan carved these runes’ – the medieval equivalent of the modern graffiti formula ‘XY was here.’

How did a medieval Viking get all the way to what is today Istanbul and was back then Constantinople, the centre of the Byzantine Empire, the most powerful metropolis on earth? Maybe Halfdan was a mercenary in the Varangian Guard. Drawn from all over Northern Europe, the Varangian Guard were an elite army unit serving as personal body guards of the Byzantine Emperor. The Byzantine Emperors felt safer with foreigners as body guards who had no local loyalties. Little is known about the motivations of the young men who left Northern Europe to serve far from home in present-day Turkey but I imagine the usual mixture of lack of opportunities at home and the lure of the metropolis – a lure so powerful that medieval Constantinople drew migrants from all across the known world to this multilingual and multicultural city.

Evidence of contemporary ‘super-diversity:’ Chinese flier in Antwerp (Source: Blommaert&Rampton, 2011)

The Viking graffiti in Hagia Sophia reminded me of the Chinese flier in a contemporary Antwerp shop window that Jan Blommaert and Ben Rampton recently used as example to explain the scope of linguistic research under conditions of super-diversity. Arguing that the example – an ad for a room for rent – bears traces of worldwide migration flows which make language varieties and scripts globally mobile, they outline the theoretical and methodological implications of migration and globalization for contemporary sociolinguistic research. I largely agree with their conclusions but I cannot help but wonder that two qualitatively similar examples – Viking graffiti in 9^th century Constantinople and a hand-written Chinese flier in 21^st century Antwerp – have such different effects: why has sociolinguistics been oblivious to linguistic diversity through the ages and why is the recognition that linguistic diversity is fundamental to all research in language and communication relatively recent?

Why does evidence of contemporary linguistic diversity move us to re-think sociolinguistics in a way that evidence of linguistic diversity through the ages has not? I answered that question previously with reference to the position of key linguistic thinkers in monolingual environments. However, there is another answer, too, and – like the medieval Viking graffiti – it also stares you in the face here in Istanbul. That further explanation is that multilingualism has been actively expunged from the historical record.

Ottoman Turkish inscription above the gate through which Mehmed II entered the city: its Greek name is Χαρ[ι]σίου πύλη/πόρτα (‘Gate of Char[i]sius’) and its Turkish name is Edirnekapı (‘Adrianopole Gate’)

However, the transient nature of language is only part of the story why we fail to see linguistic diversity in the historical record. The other part of the story is that evidence of linguistic diversity has been systematically erased from the historical record.

This obelisk inscribed with Egyptian hieroglyphs has been part of Istanbul’s linguistic landscape since the 4th century when Emperor Theodosius had it brought in from Egypt. The pedestal with its bilingual Greek and Latin inscription was added at the same time.

When Halfdan wrote his Viking graffiti and, presumably, spoke some form of Old Norse with those of his fellow Varangians who shared his dialect, the main language of Constantinople – and the lingua franca of its diverse population – was (medieval) Greek. Latin was also widely used and then there were the languages of all the city’s migrants and visitors. Christian Constantinople was a hugely multilingual place.

The city’s linguistic make-up changed on May 29, 1453 when Mehmed II took the city: not only did the Christian city become a Muslim one – and the Hagia Sophia church a mosque – the city’s dominant languages also changed from Greek and Latin to Arabic, Persian and Turkish.

What did not change was the fact of the city’s multilingualism: Arabic was the language of prayer and religion, Persian was the language of the court and Turkish was the language of the troops. Greek found itself as the language of a now down-trodden and subjected population and, as before, there were many other languages spoken by the city’s diverse inhabitants: Armenian, Hungarian, Italian, Ladino, Russian and Serbian would have been particularly prominent.

The Turkish that came to predominate over the centuries as Istanbul’s lingua franca was itself a highly heteroglossic language. Ottoman Turkish was inflected particularly by Arabic and Persian but also by all the other languages of this great melting-pot city.

Arabic calligraphy in Hagia Sophia: Quranic verse inscribed in the dome

The city’s multilingualism and the multilingual character of Turkish officially came to an end with the proclamation of the Turkish Republic in 1923. The new Turkey wanted to sever its links with its Ottoman and ‘Eastern’ past and wanted to become modern and European. The multilingual laissez-faire of the past was now seen as decidedly ‘backward’ and ‘Eastern.’ Languages other than Turkish started to be repressed, with Kurdish as the most well-known victim of the new repression of linguistic diversity by the state. Not only was Turkey going to have only one language – Turkish – but that language was going to be ‘modernized,’ i.e. rid of the traces of other languages, particularly linguistic traces associated with ‘the East,’ i.e. Arabic and Persian.

The most well-known aspect of the Turkish language reform is the abolition of the Arabic script and its replacement with the Latin script. In one fell sweep, modern Turks lost access to their written historical record. Another target of the language reformers was Arabic and Persian vocabulary. Such words were replaced with ‘Turkish’ ones or loans from ‘modern’ European languages.

The futility of this undertaking – even if lost on everyone but the philologist – is nicely encapsulated by the word for ‘city’: Ottoman Turkish used ‘شهر‎ şehir.’ Because of its obvious association with Persian ‘شهر‎  šahr’ the language reformers saw no place for it in ‘Modern’ Turkish and cast around for a ‘pure’ Turkish word. They found it in the ancient ‘kent.’ The irony is that ‘kent’ is iself a much older loanword from Sogdian, the lingua franca of Central Asia before the Islamic Conquest.

Multilingualism has made a powerful comeback thanks to the tourism economy: this restaurant menu sports entries in 10 languages. And, no, this particular dish doesn’t sound appealling in any of them …

The reform was “a catastrophic success,” as the Turkologist Geoffrey Lewis has called it. As a result, most contemporary Turkish speakers are cut off from their linguistic and cultural heritage predating the 1930s. A famous – and also ironic – example of the monolingualization of Turkish is the fact that a major 1927 speech by Atatürk, the founder of the Turkish Republic, has had to be ‘translated’ repeatedly into contemporary Turkish so as to remain comprehensible to contemporary Turks.

In Istanbul, as elsewhere, contemporary examples of ‘super-diversity’ – the Russian ‘Sale’ signs in the shop windows, the tourist communications in all the languages of countries with strong currencies, the handwritten Arabic ‘for rent’ signs, the Kurdish music stalls – are impossible to ignore. By contrast, the fact that super-diversity has been a characteristic of Istanbul-Constantinople-Byzantium since time immemorial is easy to overlook.

Monolingualism and the Turkish language – just as all other standardized languages – are invented traditions. Diversity is, in fact, the normal human experience, as the anthropologist Ward Goodenough, who passed away last weekend, pointed out back in 1976. A research agenda that takes linguistic diversity as the basis of sociolinguistic inquiry must also include the hidden histories of linguistic diversity and modernity’s attempts to erase diversity.

Jan Blommaert, & Ben Rampton (2011). Language and superdiversity Diversities, 13 (2)
Goodenough, W. (1976). MULTICULTURALISM AS THE NORMAL HUMAN EXPERIENCE Council on Anthropology and Education Quarterly, 7 (4), 4-7 DOI: 10.1525/aeq.1976.7.4.05x1652n