A health clinic in a large European city uses an interpreter request form with a dropdown menu. A patient arrives whose passport is Portuguese and whose parents came from Guinea-Bissau. She grew up partly in Lisbon and partly in Leicester. She speaks Kriol with family, Portuguese with officials, English at work, and a little Wolof with neighbours. The receptionist selects 'Portuguese.' The interpreter arrives and the consultation proceeds. But the patient describes her symptoms in terms the interpreter only half-recognises, and the most important details surface in fragments of English.

Nothing went dramatically wrong. Yet the encounter exposes an assumption built into much public administration: that each person belongs to one nationality, one ethnic group, and one language, and that these three line up neatly. For decades, especially in systems designed around post-war migration, this was a workable approximation. In many contemporary cities, it no longer is.

The sociologist Steven Vertovec named this condition superdiversity. He meant not simply more diversity, but a diversification of diversity itself. Linguists have since adopted the term to describe cities where repertoires, not languages, are the meaningful unit. In these cities, the categories used to count, serve, and govern people consistently miss how they actually communicate. This article examines what the framework offers, what urban linguistic life looks like through its lens, and what it demands of institutions that still think in dropdown menus.

Steven Vertovec introduced the term in 2007 while studying migration to Britain. He observed that since the early 1990s, newcomers had been arriving from far more countries, in smaller numbers, and through a widening range of channels. These included asylum claims, student visas, family reunification, European free movement, temporary work schemes, and irregular routes. Crucially, how people arrived shaped their legal rights, access to work, and settlement patterns as much as where they came from. Vertovec called this a 'diversification of diversity.'

Earlier frameworks were built for a different pattern of migration. Post-war multicultural policy imagined large, relatively settled communities, such as South Asian Britons, Turkish Germans, and Maghrebi French. Each was associated with a heritage language and a network of community institutions. Joshua Fishman's work on language maintenance and shift fits this model well. It treats the community as the unit of analysis. It asks whether a language survives across generations, which domains of life it occupies, and which institutions sustain it.

Superdiversity unsettles that unit. Two Somali speakers in Rotterdam might be a Dutch-born citizen and a recently arrived asylum seeker from Ethiopia's Somali Region, with almost nothing in common administratively. Linguists such as Jan Blommaert and Ben Rampton argued that this calls for a shift from languages to repertoires. A repertoire is the uneven bundle of registers, accents, scripts, and partial competences that a person acquires over a lifetime. It records a life trajectory rather than membership in an ethnic group.

The concept has drawn serious criticism. Aneta Pavlenko argued that it rebrands long-standing phenomena, pointing to historic port cities and empires that were always intensely multilingual. Scholars in the Global South note that Lagos, Mumbai, and Johannesburg hardly needed a European term to describe linguistic complexity. Nelson Flores and Jonathan Rosa, working within raciolinguistics, warn of a further risk. Celebrating fluid complexity can obscure how race still determines whose language mixing is heard as creative and whose is heard as deficient.

These critiques are useful because they clarify what the framework can legitimately claim. Superdiversity works best not as a statement that diversity is new, but as a diagnosis of institutional mismatch. The categories that public systems inherited, one person, one ethnicity, one language, were designed for a particular era of migration. Read this way, the concept is less a label for a city than a lens. It helps us notice where official descriptions and lived communication have drifted apart.

Takeaway

Superdiversity is less a claim that cities are more mixed than ever and more a warning that the categories we use to describe them have quietly expired.

The linguist Einar Haugen proposed thinking about languages ecologically. In his view, they are not isolated systems but resources that exist in relationship with their speakers, their environments, and one another. In superdiverse neighbourhoods, this ecology is organised less around community territories than around sites of encounter. Markets, barbershops, phone repair shops, school playgrounds, and bus routes become places where people with very different repertoires routinely need to get things done together. The relevant question shifts from 'which communities live here?' to 'what communicative work happens here, and how?'

Researchers have developed several overlapping terms for what happens in these spaces. Emi Otsuji and Alastair Pennycook describe metrolingualism, the everyday mixing of linguistic resources in places like Sydney's produce markets. Ofelia García and Li Wei use translanguaging for speakers who draw on their full repertoire rather than switching between separate languages. Suresh Canagarajah's work on translingual practice highlights negotiation strategies, such as letting unclear words pass and building shared meaning gradually. His work suggests that competence is achieved in interaction, not simply possessed.

Young people are often at the leading edge. Multicultural London English, Kiezdeutsch in Berlin, straattaal in Dutch cities, and similar varieties in Stockholm and Copenhagen draw features from Caribbean creoles, Arabic, Turkish, Yoruba, and local working-class speech. Notably, young people of many backgrounds speak them, including the children of long-established residents. Rampton calls them contemporary urban vernaculars. They are not imperfect versions of the national language. They are new local norms with their own systematic features.

The city's surfaces record these dynamics too. Linguistic landscape research treats shop signs, posters, and handwritten notices as evidence of who lives, trades, and is addressed in a space. Blommaert's long-term study of his own neighbourhood in Antwerp showed signs layered like sediment. A grocery might operate under a former bakery's lettering, beside church notices in African French and municipal announcements in standard Dutch. Each sign speaks to different audiences, reflects different histories, and signals a different degree of permanence.

None of this amounts to a harmonious mosaic. Urban linguistic ecologies are stratified. Some resources, such as standard English or educated French, carry value across contexts, while others are stigmatised or invisible to institutions. Classrooms frequently police youth vernaculars as 'slang' or as a sign of deficiency. Blommaert's notion of scales captures this. A repertoire that works well in a local market may lose its value entirely at a job interview or a benefits office. People can move freely without their linguistic capital moving with them.

Takeaway

In a superdiverse city, the question 'what language do you speak?' matters less than 'what can you do with the resources you have, and where do they count?'

Public administration tends to manage language through lists. Censuses ask for a main language. Interpreting services keep registries organised by language name. Councils translate leaflets into the 'top five' or 'top ten' languages in their area. In superdiverse conditions, these lists run into a long-tail problem. A few languages have large numbers of speakers, followed by dozens or hundreds with small ones. Prioritising the largest groups is administratively rational, yet it systematically excludes those who are already least visible.

The data problem runs deeper than coverage. A single 'main language' question flattens repertoires. A resident comfortable with spoken English but not with written bureaucratic English may be recorded as needing no support at all. Language names also mislead. 'Arabic' covers varieties that are not always mutually intelligible in a stressful conversation, and 'Kurdish' may mean Kurmanji or Sorani. Matching interpreters by nationality compounds the error. More useful questions ask which language someone prefers for medical, legal, or written communication specifically.

Where formal systems fall short, informal arrangements fill the gap. Children interpret for their parents at doctors' appointments and housing offices, a practice known as language brokering. Neighbours and volunteers step in. Staff increasingly use machine translation at counters and in emails. These practices keep services running, but they carry real costs. Confidentiality can be breached, errors go unchecked, minors take on inappropriate burdens, and accountability becomes unclear. Policy often swings between quietly relying on such arrangements and banning them without offering alternatives.

There is also an ideological tension. Many integration frameworks treat learning the national language as both a duty and a precondition for full participation. Some even frame translation as a disincentive to learn. Superdiverse reality complicates this, because people may be actively learning while still needing support today. Fishman's emphasis on community institutions also meets a new obstacle. Funding schemes often expect a single representative community organisation. Small, internally divided, recently arrived groups rarely fit that template.

Adaptive approaches share a common principle: shifting the burden of communication from the individual to the institution. Practical measures include regionally pooled telephone and video interpreting for rarer languages, and plain-language, visual design that reduces reliance on translation. They also include accredited training for community interpreters and formal recognition, including pay, for multilingual frontline staff whose skills are often used but unrewarded. Training staff to check understanding, rephrase, and slow down matters as much as any translation budget. Participatory language mapping with residents can keep the data current.

Takeaway

An institution's language policy is tested not by the languages it lists but by what happens to the person whose language is not on the list.

Superdiversity is most valuable not as a description of how mixed a city has become, but as a prompt to audit the categories institutions use to see residents. Where those categories assume one person, one group, one language, they will consistently misread the people they are meant to serve.

Several strategic directions follow. Collect data on repertoires and context-specific preferences rather than single main languages. Budget explicitly for the long tail instead of treating rarer languages as exceptions. Recognise urban vernaculars and multilingual staff as assets. Protect children from routine brokering duties. And involve residents in mapping their own linguistic environments, since they understand them better than any form can.

Return to the clinic and its dropdown menu. The meaningful question was never which box to tick, but what would allow this particular person to be understood. Cities that learn to ask that question will be better equipped for the linguistic complexity that is already here.