Is Kurdish a Language or a Dialect? The Clear Answer of Linguistics

At a Glance

Kurdish is a language in its own right. It belongs to the Iranic branch of the Indo-European family, with its own grammar, sound system, and written literature. Among linguists this is not in dispute.
"Iranic" is the name of a language family, not the modern state of Iran. English is a Germanic language, yet the English are not Germans; Kurdish is an Iranic language, and Kurds did not descend from the Iranian state.
Kurdish and Persian are sister languages, not parent and child. Dr. Mehrdad R. Izady measures the distance in one image: Kurdish is as close to Persian as German is to Danish. Kin, but two separate houses.
The main branches of Kurdish are Kurmancî, Soranî, and Southern Kurdish (Kelhûrî, Feylî). The place of Zazakî and Goranî on the family tree is debated among linguists; the identity of the people who speak them is a separate matter, and it rests on self-definition.
Kurdish has two grammatical genders, case inflection, and an ergative construction. Turkish has none of these; the claim that "Kurdish is broken Turkish" cannot stand for a minute in front of linguistics.
Kurdish has at least three hundred years of unbroken written literature: from the divan of Melayê Cizîrî to Ehmedê Xanî's Mem û Zîn, from the first newspaper in 1898 to the novel of today.
"Language or dialect" is usually a political question, not a scientific one. The linguist Max Weinreich passed on the famous line that sums it up: "A language is a dialect with an army and a navy."
The number of Kurdish speakers is not known for certain; estimates put the population between 30 and 45 million. Even the absence of a firm count is a result of the language's suppressed history.
Kurdish is written today in two alphabets: Latin (mostly for Kurmancî) and an Arabic-based script (mostly for Soranî). This doubling is the mark, in the language itself, of a people split across four states.

A child comes home. The grandmother greets him in Kurmancî: "Rojbaş, lawê min." Good day, my son. The child answers, because this is the language of the house. His lullaby is in it. His fairy tale, his scolding, his love.

The next morning he is at school. The teacher writes on the board. The language of the school is another one, and the rule is plain: the language of the house is not spoken here. Over the years the child learns something no one states outright. It seeps from the walls: the language of the house is a "small" thing. One day a teacher says it plainly. "That isn't even a language."

This scene played out millions of times in the twentieth century. And most of those children, once grown, asked the same question. Was my grandmother's language really not a language?

This article is the answer linguistics gives. The answer is short: it is a language. What is long is the evidence behind that short answer.

Who Decides Between "Language" and "Dialect"?

Let me put the concepts on the table first. A dialect is one of the regional branches of a language. The English of London and the English of New York are dialects of one tongue; their speakers understand each other.

So who decides whether two ways of speaking are separate languages or dialects of one? Linguistics has a main test: mutual intelligibility. If two people, each speaking their mother tongue, understand each other, they are inside one language. If they do not, they are in separate languages. To this you add a comparison of structure: sound system, grammar, vocabulary.

There is also a political face to this, and no need to hide it. Weinreich's line says it all: "A language is a dialect with an army and a navy." The standing of a "language" is most often granted by state power; the tongue of the stateless is belittled as a "dialect." Swedish and Norwegian are very close, yet because there are two states there are two languages. Kurdish and Turkish come from entirely separate families, yet because Kurdish had no state it was stamped a "dialect." That gap between science and power is the heart of our subject.

Whose Child Is Kurdish?

We sort languages into families because languages, too, are kin. So where does Kurdish sit? Its family tree runs like this. At the top stands the Indo-European family: the vast family of English, German, Russian, Greek, Persian, and Hindi. One branch of it is the Indo-Iranian branch; one arm of that is the Iranic languages. Kurdish sits in the northwestern group of that arm. The documents and the linguistic literature agree on this placement; from the British linguist D. N. MacKenzie's classic 1961 study to today's research, the line is the same.

Here a vital warning is needed. The word "Iranic" is the name of a language family, not of the modern state of Iran. Let me fix the distinction with a comparison. English, as a language family, is a Germanic language; yet no one counts the English as Germans, or sees England as part of Germany. In the same way Kurdish is an Iranic language; this does not mean Kurds descended from Iran, or that Kurdish is a branch of Persian. Kurdish and Persian are sister languages; one is not the child of the other. Both reached today by separate roads out of an old Iranic pool. Dr. Izady, the backbone author of Kurdish studies, gives the measure in a single image: Kurdish is as close to Persian as German is to Danish. Kin, but two separate houses.

Turkish comes from an entirely different family: the Turkic family. So the kinship between Kurdish and Turkish is the kinship between English and Japanese: none. Centuries of neighbourhood naturally bred an exchange of words; but borrowing words is not kinship. Turkish holds thousands of Arabic words, and no one calls Turkish an Arabic tongue.

Kurdish's Own House: Its Branches

Kurdish is spoken across a wide geography, and like every large language it splits into branches within itself.

The most crowded branch is Kurmancî, spoken in the north: along the Turkey, Syria, and Caucasus line, and in the northern strips of Iraq and Iran. The second large branch is Soranî, common in the centre of the Kurdistan Region of Iraq and in the Kurdish provinces of Iran. In the south are branches such as Kelhûrî and Feylî, together called Southern Kurdish.

Then there are Zazakî and Goranî, and here honesty is required: the scholarship splits in two. Dr. Izady gathers the Kurdish language world into two main blocks: on one side the Kurmancî block (Northern Kurmancî, Soranî, and Kelhûrî branches), on the other the Pehlewanî block (Dimilî, that is Zazakî, plus Goranî and Lekî). In this reading Zazakî and Goranî stand not outside Kurdish but in its oldest layer; Goranî's written literature, reaching back a thousand years, is that block's title deed. The opposing line follows MacKenzie: these branches are not inside Kurdish but a separate arm of the same northwestern Iranic family. Scholars are divided here, and the debate goes on.

But let me draw one line thick. A place on the family tree is one thing; identity is another. As the researcher Malmisanij has documented, most Zazas in Turkey have long counted themselves Kurds; even the names they give themselves, Kird and Kirmanc, carry this. On identity the last word belongs not to the measuring stick but to the person. Linguistics argues the family tree; people say who they are themselves.

How well do the branches understand each other? Between Kurmancî and Soranî, understanding is hard; the grammars differ seriously. Does this take Kurdish out of the category of a language? No. Between the northern dialects of German and Swiss German, understanding is also hard; between the Moroccan and Iraqi branches of Arabic it is nearly impossible. A wide geography, a deep history, and the borders of four separate states: dialect differences are the natural result.

The Claim That "Kurdish Has No Grammar": What Does the Record Say?

In the twentieth century much was said about Kurdish: it has no vocabulary, it has no grammar, it is mountain Turkish. What these claims had in common was that the people making them did not know Kurdish. So let me open the record. What does it actually show?

In Kurdish, in the Kurmancî branch, nouns have two grammatical genders: masculine and feminine. "Çem" (river) is masculine, "çiya" (mountain) is masculine, but "roj" (sun, day) is feminine. Anyone who knows French or German recognizes this system. Turkish has nothing of it.

Kurdish has case inflection; nouns change shape according to their job in the sentence. And one of Kurdish's most interesting subtleties is the ergative construction. Let the name not frighten you; the idea is simple. In Kurdish, in the past tense, the subject of a transitive verb takes a special form, and the weight of the sentence shifts to the object. To say "I ate the apple," Kurmancî puts "I" into another mould: "Min sêv xwar." This construction exists in only a small share of the world's languages; it places Kurdish in the same elegant club as Basque and Georgian. The record is clear: the grammar of the language called grammarless is rich enough to fill article after article for linguists. The 2014 survey by Geoffrey Haig and Ergin Öpengin maps that literature.

A language shows it is a language through its vocabulary, too. Kurdish carries the full richness of terms for mountain country, herding, farming, and a strong oral literature. There are separate words for snow, for naming a sheep by its age and colour, for degrees of kinship. The vocabulary of peoples is the mirror of the life they live.

A Bookshelf Three Centuries Long

To those who say "a dialect has no literature," Kurdish's answer is a bookshelf. But how far back does it reach?

At the head of the shelf stands the divan of Melayê Cizîrî: poems of mysticism and love, written at the turn of the sixteenth into the seventeenth century in Cizre. Beside him Feqiyê Teyran, the poet who spoke in the language of birds. Then 1694: Ehmedê Xanî's Mem û Zîn, the masterpiece of Kurdish. We told Xanî's story in a separate article; here one sentence is enough. Xanî recorded in his own hand that he wrote it so no one could say "the Kurds have no book."

The shelf widened with the press in the nineteenth and twentieth centuries. In 1898 the first Kurdish newspaper appeared in Cairo. In the 1930s, in Damascus, the journal Hawar established Kurmancî in the Latin alphabet. Radio Yerevan and Radio Baghdad moved Kurdish into a sound archive. The Soranî branch became the language of the press, of poetry, and of administration in Silêmanî and Hewlêr. Today Kurdish novels are written, taught at university, spoken on television, and, as these lines show, made the subject of scholarly research.

Three hundred years of unbroken written tradition. How many "dialects" can say that?

How Many People Speak It?

The honest answer: no one knows for certain. Why can no one give a firm number? Most of the countries where Kurds live never ran censuses that counted mother tongue properly; in some periods, even saying you spoke Kurdish took courage. Estimates therefore roam a wide range: the figures spoken for the Kurdish population run between 30 and 45 million. The number of speakers is lower than the population, because assimilation, the seizing of a people's language and identity, tore millions of Kurds from their mother tongue across generations. In the big cities, grandchildren who cannot speak a word of Kurdish with their grandmother are the living record of that process.

In the international language catalogues, Kurdish is registered under an umbrella code together with its branches. Let me add this: code tables are administrative tools, not the referees of identity. The worth of a language is set neither by a code nor by a border. The worth of a language is set by those who keep it alive.

Why Is the Question Still Asked?

If linguistics gave its answer a hundred years ago, why is "is Kurdish a language" still in circulation?

Because the question was never a question of linguistics. To declare a language a "dialect" is the first step toward declaring the people who speak it a "sub-group." When the language goes, folklore goes after it, then history, then the name. This is why in the twentieth century Kurdish was not only banned; it was treated as nonexistent, its name changed, gathered from the mouths of children. And this is why speaking, writing, and reading Kurdish was counted, for generations, as an act of existence.

To the child who asked about his grandmother's language, we can now answer. That language comes from the Iranic branch of the Indo-European family. It has two genders, an ergative construction, a bookshelf three centuries long, and tens of millions who keep it alive. Whatever was said at school, the lullaby was in a language. And languages survive in lullabies.

Frequently Asked Questions

Which language family does Kurdish belong to? To the northwestern group of the Iranic branch of the Indo-European family. It is a sister language of Persian, not a branch of it. It has no kinship with Turkish; Turkish belongs to a separate family entirely.

What are the dialects of Kurdish? The main branches are Kurmancî, Soranî, and Southern Kurdish (Kelhûrî, Feylî). The place of Zazakî and Goranî on the family tree is debated: Dr. Izady counts them as Kurdish's Pehlewanî block, while the MacKenzie line classes them as a separate arm. Most of their speakers count themselves Kurds, and in identity self-definition is decisive.

Do Kurdish and Persian understand each other? No. Both are Iranic languages, but like Spanish and Italian they are separate languages; despite a common root they developed separate grammar and vocabulary.

How long has Kurdish had a written literature? At least since the end of the sixteenth century: the divan of Melayê Cizîrî, the poems of Feqiyê Teyran, Ehmedê Xanî's Mem û Zîn in 1694, the first newspaper in 1898, the journal Hawar in 1932, and the novel, press, and academy of today.

In which alphabet is Kurdish written? In two: mostly the Latin alphabet for Kurmancî (the Hawar alphabet), and mostly an Arabic-based script for Soranî. This doubling is the mark, in the language, of a people split across four states.

Sources and Further Reading

Primary sources (from the Bedel Boseli Collection):

Dr. Mehrdad R. Izady, The Kurds: A Concise Handbook (4th edition), the language chapter; the two-block classification (Kurmancî and Pehlewanî) and the German-Danish comparison come from here.
Malmisanij, Kırd, Kırmanc, Dımıli veya Zaza Kürtleri; on the self-naming and identity of the Zazas.
Michael L. Chyet, Kurdish-English Dictionary; for vocabulary and definitions.

Social Media Summaries

The following summaries are ready to share; copy and use them.

"A language is a dialect with an army and a navy." Swedish and Norwegian count as two languages because there are two states; Kurdish, stateless, was stamped a "dialect." Science says otherwise: bedelboseli.com/en/is-kurdish-a-language-or-a-dialect
Kurdish is an "Iranic" language, but that is a family name: English is "Germanic," and the English are not Germans. Kurdish is not the child of Persian but its sister: bedelboseli.com/en/is-kurdish-a-language-or-a-dialect
The language once called "grammarless" has two genders, case inflection, and an ergative construction found in few tongues on earth. Kurdish sits in the same elegant club as Basque and Georgian: bedelboseli.com/en/is-kurdish-a-language-or-a-dialect
From Cizîrî's divan to Xanî's Mem û Zîn, from the newspaper of 1898 to the novel of today: Kurdish has a bookshelf three centuries long and unbroken. How many "dialects" can say that? bedelboseli.com/en/is-kurdish-a-language-or-a-dialect
To declare a language a "dialect" is the first step to declaring its people a "sub-group." That is why the question was never linguistic; it was always political. The answer stays the same: bedelboseli.com/en/is-kurdish-a-language-or-a-dialect