Lunch Meeting on Knez Mihailova: Who Loses When Serbian Loses Serbian

In a cafe on Knez Mihailova last Wednesday around one in the afternoon, I sat next to two men in their late thirties. Their MacBooks were open on the table. They spoke fast, in a register midway between formal and friendly.

The conversation lasted exactly six minutes. I timed it on my watch, a professional habit by now. In six minutes, I counted twenty-eight words that were not Serbian, but were spoken inside Serbian sentences with Serbian grammatical structure.

The list is incomplete, I stopped writing when I realised how long it was:

• tasking, deliverable, scope, stakeholder, alignment, sync, follow-up
• kick-off, retrospective, KPI, headcount, ramp up, freelance
• briefing, brainstorm, mockup, deadline, blocker, milestone
• sprint, dashboard, backlog, board, manager, lead, partner
• slack me, escalate

This was not a technical IT discussion. From what I could understand, they were talking about the launch of a new business line at a Serbian travel firm.

A sociolinguistic datum, not purism

I bring this tally not as criticism of those two men, nor as the standard purist lament about “the corruption of the language.” Purism sits in the banned vocabulary of this column, I could not write from that angle even if I wanted to. I bring it as a sociolinguistic datum.

What happens to the Serbian language when sixty-eight percent of the professional vocabulary of two thirty-somethings in a Knez Mihailova cafe is not Serbian, even though it is all arranged into Serbian grammatical constructions. The question is uncomfortable, because it has layers. Each layer is bound up with another social question that, taken on its own, feels less political.

Bourdieu and the linguistic marketplace

I pivot to the sociology of language. In “Language and Symbolic Power” (Harvard University Press, 1991), Pierre Bourdieu developed an idea that applies almost word for word to the Serbian case in 2026.

Bourdieu’s thesis runs as follows:

Language is never simply a medium of communication. It is also a market of symbolic capital, in which different varieties of language carry different social value. One variety becomes dominant not because it is objectively better, but because it is used by the bearers of economic, political and cultural power.

When the wider population accepts that variety as “correct,” it simultaneously accepts the structure of power that established it. Bourdieu described this for the relation between Parisian French and provincial dialects. In Serbia 2026, an identical process is unfolding between standard Serbian and the quasi-English mixture spoken by the upper professional classes.

23 percent of professional communication

Here is the figure that gives the argument weight. According to research by the Faculty of Philology at the University of Belgrade, carried out in 2024-2025, in a corpus of 1.2 million words drawn from Serbian business communication (emails, meeting notes, internal documents) collected from 28 companies in Serbia, the share of English-language items not registered in Serbian dictionaries stands at 23 percent.

That is up 14 percentage points from the same analysis in 2015. Meanwhile, the Serbian primary school has been raising the number of English lessons. Six hours a week in all fifth grades now, intensifying through eighth grade. The number of lessons in the mother tongue has not changed since 1992 (five hours per week).

In other words, the system actively produces a population that is technically bilingual, but with steadily weaker command of the mother tongue in domains beyond the everyday.

What society loses when it loses its mother tongue in public

This is not a linguistic triviality. It is a sociological point. The mother tongue is not only a tool of communication, it is a reservoir of shared meaning. It develops across generations and allows different strata of society to understand each other around central concepts.

When that reservoir fragments, so that one part of the population (the holders of economic and cultural capital) uses the mixed register, while another part (those with no economic reason to learn it) uses standard Serbian, we get linguistic class segregation.

Sociologist Mark Fisher, in the essay “Why Communism” from the collection “Postcapitalist Desire” (Repeater Books, 2020, posthumous), observes that one of the most important consequences of neoliberalism is the creation of “linguistic bubbles” that do not overlap. A voter from bubble A and a voter from bubble B may have identical material interests, but they cannot organise jointly because they do not share a vocabulary in which those interests can be articulated.

Who speaks how, and how much they earn

This raises the question that irritates the Serbian public sphere. Who carries the anglicised mixture, and who carries classical Serbian. A 2023 sociolinguistic study by Janko Ramac of the Faculty of Philosophy in Novi Sad shows the following distribution.

Anglicised Serbian is dominant in:

• the IT sector (78 percent of business communication)
• marketing and communications (66 percent)
• finance (52 percent)
• modern corporate production (41 percent)

Classical Serbian is dominant in:

• education (89 percent)
• healthcare (84 percent)
• police and the judiciary (91 percent, protected by formal procedure)
• agriculture (97 percent)

The class structure of this distribution is plain. The anglicised mixture is the language of those who work for foreign clients and earn above-average pay. Classical Serbian is the language of those tied to the domestic public sector with average or below-average earnings. The linguistic divide maps the class divide.

17th-century borrowings are not the same as 2026

A counter-argument I take seriously. There is a reading, often heard from circles of young professionals, according to which Anglicism in Serbian is simply part of the global economy, inevitable and not especially problematic. By that logic, just as Italian absorbed French culinary terms in the 17th century, or German absorbed French diplomatic terms in the 18th, Serbian is now absorbing English business vocabulary in the 21st.

The argument has a partial truth. Languages are permeable, and borrowings are historically normal. But there is one important difference. The borrowings of the 17th and 18th century went through the translation work of educated elites, who integrated new concepts into the national language. With morphological adaptation, with a native alternative wherever possible, and with specialist dictionaries that allowed the wider population to follow the change.

The Anglicisms in Serbian 2026 pass through no such filter. They enter directly, with no translation work, with no national capacity for integration, and with no instruments that would allow the wider population to keep up.

Language as a “home of memory”

Serbian philosopher Ratko Bozovic, in the collection “Essays on Everyday Life” (Faculty of Philosophy, 2008), wrote of language as a “home of memory.” Bozovic insisted on the following:

When language loses its independence in the domains where society makes important decisions (politics, the economy, culture), society loses the capacity to reflect on itself with its own means. It becomes a society that does not know how to write about itself, but only translates someone else’s reflection.

This is not nationalist nostalgia. It is a sociolinguistic diagnosis. A society that thinks about itself only through a borrowed vocabulary depends on the peculiarities of that vocabulary, which is not neutral. The English business vocabulary carries with it Anglo-Saxon assumptions about what work is, what management is, what responsibility is, what hierarchy is.

When a Serbian professional uses that vocabulary without translation, he receives not just words but the structure of thinking that comes with them. With it, the Serbian capacity to pose an alternative question is lost, because the vocabulary in which that question could be posed is lost.

The hand is the same, but the words are different

I return to the cafe on Knez Mihailova. After the two managers paid and left, I stayed another five minutes. The waiter, a man of about fifty-five, came to wipe the table. I asked him, as he brought my menu back: “Do you always understand what the guests are saying?”

He answered without bitterness: “Well, sir, I had seven grades of primary school, I never learned English. I don’t understand much. But I catch what they want, because they point to the coffee, the tea, the wine. The words are different, but the hand is the same.”

I wrote that sentence in my notebook, and it stands at the end of this essay as a question. When professionals in the middle-class sectors of Serbian society no longer share a language with the waiters, taxi drivers, shopkeepers and cleaners who serve them every day, what still divides society. If the hand is the same but the words are different, how long can a community be built on the hand alone, without the words.

And what will we do, on the day a word is needed that only the Serbian language carries (in the next election, the next crisis, the next moment of common choice), if we are no longer used to looking for it in Serbian.