When Does a Dialect Become a Language?
A popular linguistics quip observes that a language is a dialect with an army and a navy. It is tragically common for national disputes (by which we ultimately mean linguistic ones) to put that theory to the test. I have argued elsewhere that, for example, language has sparked almost every European war since 1870. But, of course, languages are not entirely defined by politics (nor by politics by other means). Languages are defined and redefined more in schools, markets, and popular music than on the battlefield. So, when it comes to this fundamental distinction, where do we draw the line?
It is helpful to keep in mind that language drift is a natural phenomenon. As trade and literacy collapsed with the fall of the Roman Empire, what had been regional accents of Latin diverged so starkly that, by the 8th Century, peasants in France could no longer understand priests from Italy. Language helped define the nations of Europe, rather than the other way around.
On the other hand, China never permanently shattered. Regional dialects of Chinese did diverge, in some cases as much as different Romance languages. But in the minds of everyone, they remained dialects of the one official language. Linguistic unity has only truly been realised in recent times, but, with the exception of Cantonese in Hong Kong and Macao, no “dialect” has ever held the status of an official language.
Thus, when it comes to language, politics can be downstream of culture, or vice versa. In the 20th Century, leaders of nationalist and anti-colonial struggles recognised the central role that language plays in people’s identity. For the first time, they deliberately sought to shape it.
In 1946, Hindustani was one of the biggest languages on earth, spoken by hundreds of millions of people from the Indus to the Ganges. One year later, it did not exist, because the new nations of India and Pakistan wanted to define Hindi and Urdu as separate languages. It mattered not that people on either side of the border could, and still can, easily understand one another. They drew on existing regional differences in accent and vocabulary, some grammar quirks, and, most importantly, writing, to create two separate tongues for two separate nations.
Writing is often the key factor in dialect/language disputes. There are several reasons for this. One is that it is pretty easy to control, unlike people’s everyday speech. Another is that a writing system often signals a religious or political bearing — Urdu uses the Persian alphabet, while Hindi uses the Indian Devanagari script. In Europe, Slavs who are Catholic (e.g., Czechs, Croats, Poles) favour the Latin alphabet, while their Orthodox cousins (e.g., Russians, Serbs, Ukrainians) use Cyrillic.
Perhaps most importantly, people just tend to think of writing as the official version of a language. From a linguistics perspective, this is a fallacy: any writing system, even the best, is just an approximation of a spoken tongue. But if someone sees all the signs here written in one script, and the ones over there in some totally alien one, he will be inclined to believe that he really is speaking a different language.
Creating a national orthography can be a key step in nation-building. In the 1880s, under pressure from British expansion, Dutch-descended Boers sought to emphasise their deep roots in their continent by redefining their language from a dialect of Dutch to a new language called Afrikaans (the Anglo-dominated South African government did not recognise it as such until 1925). Of course, the use of local lingo and grammatical constructions did not, at first, change how Afrikaners actually spoke. But over time, it legitimised Afrikaner-isms from dialectal laziness to features of a separate language. Afrikaans did draft farther and farther from standard Dutch.
An analogous process has occurred in Haiti. Since Creole became recognised as the world’s newest national language, it has drifted so far from French as to be barely recognisable. But in more developed countries, the shift is usually not so dramatic.
When the royal personal union between Denmark and Norway dissolved in 1905, Norwegians were keen to establish their national distinctiveness. But to this day, they cannot agree whether to write using Bokmål spellings, which are closer to Danish and the language of southern port cities, or those of Nynorsk, which are more “authentically” Norwegian. Meanwhile, Norwegians can still pretty easily understand spoken Danish.
The publication of Noah Webster’s American English dictionary in 1828 was a key moment in establishing an American national identity. But, as in Norway and the Indian Subcontinent, political efforts to divide people can only go so far. Americans and Britons still speak the same language, even if they insist on spelling a few words differently. When the state of Illinois in 1923 declared that their official language was American, they were widely ridiculed (interestingly enough, the move was driven by Irish and Jewish politicians upset with British policy in Northern Ireland and Palestine). Scottish nationalists’ efforts to promulgate Scots as a standard written language have largely fallen flat.
Language disputes among people who speak the same language are not always so civilised as spats over spelling. From a linguistic perspective, Croats, Serbs, and Bosnians are one people, but that did not stop them from trying to wipe one another out just thirty years ago. Language engineering for nationalist purposes, as explored above, is a modern phenomenon, but organic xenophobia has deep roots. In a rebellion in the 1520s, peasants in Friesland (in the northern Netherlands) strung their closely-related Dutch neighbours up if they were unable to speak proper Frisian. To some extent, all the wars in history between (and within) France, Italy, Portugal, and Spain can be traced back to diverging dialects of Latin.
The process can also go in reverse, with similarly varying success. In 1789, fewer than half of Frenchmen spoke standard French. Occitan was as much a foreign language as Spanish. Thanks to vigorous language policies, some of them verging on the psychotic, today, all of them do. You can easily get around anywhere in Germany in standard German; Spain and Italy run only slightly behind. China is working on it. By contrast, 1930s efforts to make western Ukrainians speak standard Russian, or to create a new Czechoslovak language, ran aground.
So, what does distinguish a language from a mere dialect? On paper, armies and navies are, indeed, of prime importance. But on the ground, what matters are the hearts and minds of speakers. Afrikaners have diverged very markedly from their old mother country; New Zealanders have not; and the Québécois are somewhere in between. Two centuries ago, Norwegian and Danish were clearly separate languages, while Haitians spoke French; today, the Norwegians hew much closer to their old overlords than the Haitians do.
Dialects typically diverge over time, but for many major languages, modern communication technology has reversed this trend. Thanks to Spotify, Netflix, and other apps, young Spaniards pepper their language with Latin American imports, and Singaporeans speak more and more like Mainland Chinese. Indians and Pakistanis generally think of each other as implacable rivals, but nevertheless enjoy the same music and movies; popular historical dramas like Bajirao Mastani use both registers interchangeably.
Language is at the core of every culture. The way people think about it is, naturally, influenced by myriad factors that have nothing to do with linguistic logic. If you go to Macedonia and tell a Macedonian cab driver, in Bulgarian, that he is speaking Bulgarian, he will understand you perfectly and will probably cuss you out for saying it. If you tell a Fujianese farmer, in standard Chinese, that he speaks Chinese, he may need a translator, but will agree. In language, as in other areas, armies and navies can do a lot — but they cannot really make people rational.
