Kati, the language of the Kata and the Kom

Halfmann describes the language as spoken in adjacent sections of five valley systems. In the west, Ramgal at the top of the Alingar, with more than twenty villages, and Kulam, a side valley lower down, cut off from Ramgal by Pashai-speaking country; both are in Mandol district. In the centre, Kantiwa, the Pech tributary the Kata count as their ancestral home, from which they spread in about the seventeenth century, with the villages of Kantiwa, Aspit and Mum. In the east, the whole Bashgal valley, which the maps call the Landai Sin, from Bargi Matal and the Kata villages of the upper valley down through Kamdesh and the Kom country to where it meets the Kunar, and on into the upper ends of a few Kunar side valleys such as Dungal. And across the Durand Line in Chitral, in Gobor at the head of the Lutkuh valley, in Shekhandeh above the Kalash villages of Rumbur and Bumburet, and in Langurbats and Badugal along the main Chitral river.

Not every village that was Kati-speaking still is. Strand records that Persian from the Panjshir has replaced Kata-vari in four villages of Ramgal, and that Pashto has displaced Kamviri in six Kom villages on the Kunar side. Two villages at the top of the Munjan valley in Badakhshan, Naw and Ferazen, still speak it, perhaps two hundred people, the one Nuristani outpost north of the watershed. And there are Kati speakers in Kabul, Jalalabad, Peshawar and Chitral town, and in the diaspora abroad.

The Chitral community deserves a word. Its families moved from Nuristan in the late nineteenth century, around the time of the conquest, and now number about six thousand eight hundred by Pakistan's 2018 census of the relevant villages, though the community itself claims ten thousand. They call their language Shekhani or Kataviri there, and since 2021 they have been the most active of all Kati speakers in writing it down.

Kati is one of the five Nuristani languages, which form their own branch of the Indo-Iranian family beside Iranian and Indo-Aryan. Morgenstierne made the case in 1926 and sealed it in 1945 with a paper on how the family treats an old Indo-European consonant, and Strand's account in the Encyclopaedia Iranica sets it out: where Nuristani has a dental ts, Iranian has s and Indo-Aryan sh, so that ten is duts in Kamviri, dasa in Avestan and Sanskrit. Halfmann's 2025 book argues the branch separated early from the Iranian side and only later came under the heavy Indo-Aryan influence that makes Kati, Waygali-ala, Ashkun and Tregami look in many respects like New Indo-Aryan languages.

Where Kati sits inside the family is less settled. Strand groups it with Prasun in a northern cluster, on shared sound changes, while noting that Prasun was cut off from the developments the core languages share. Glottolog, following the comparative work of Irén Hegedűs, puts Kati with Ashkun instead, in a node it calls Ashkun-Kate, and gives Prasun a branch of its own. Strand's own 2023 paper withdraws the northern group, pointing out that eighteen per cent of Prasun's vocabulary is borrowed from Kati, which was the common language of the province and made the two look closer than they are. The honest summary is that Kati is central to the family, and that its nearest relative is still being argued over.

The first book on the language was written by Colonel John Davidson of the Indian Army, who spent two years on duty in Chitral from 1898 and worked with two men from Kamdesh and the Bashgal, Sher Malik and Gul Mir, to translate about 1,741 English sentences into what he called Bashgali. His Notes on the Bashgali (Kafir) Language appeared in Calcutta in 1902, and Sten Konow turned it into a dictionary in 1913. Davidson recorded a tradition among his informants that their people had once practised reading and writing. Morgenstierne took up the language from 1926, visited the Kati of Bumburet in 1929, published Kati myths and hymns in the 1950s and mythological texts from the Kata in 1968, and left an unfinished Kati vocabulary in the National Library of Norway.

The first full grammar was written by a Kom. Qazi Ghulamullah of Kamdesh, a jurist, completed a Persian-language grammar of Kamviri of six hundred and fifty pages in 1966, now in the University of Oslo library, and a Pashto-Kamviri dictionary in 1968, held at Bern; neither was ever printed. He advised Morgenstierne, Strand and Gérard Fussman, and devised his own Arabic-script spelling with invented diacritics that nobody else adopted. He died in 2007. Strand began his fieldwork among the Kom in 1967 and has published on them for more than fifty years: the history, kinship and rank of the Kom, the Kamviri verb, an online Kamviri lexicon of some 5,500 words, a Kata-vari lexicon of Kantiwa and an etymological lexicon of the whole family. Grjunberg, working in Afghanistan from 1963 to 1967, published the Russian grammar and texts of the Kulam dialect in 1980, which the Encyclopaedia Iranica calls the best in its series.

The newest and largest work is Halfmann's A Grammatical Description of the Katë Language, a Cologne doctoral thesis of 684 pages published in 2024 and free to read online, the first comprehensive grammar of the language. He built it over three and a half years from 2019 with Abdul Nasir Mansur, a school principal and language activist from Shekhandeh in Bumburet, and from recordings, radio programmes and social media in all three dialects, with two weeks in Chitral in 2022. A Kabul writer, Sulayman Sun-Aro, printed an illustrated Kati dictionary of 321 pages in 2016, and his spelling has become the model most Kati writers follow.

The Kata are the largest Nuristani people and the one that gave the language its usual name. From Kantiwa they spread west into Ramgal and Kulam and east over the passes into the upper Bashgal, a movement Alberto Cacopardo dates to about the seventeenth century. The Kom hold the lower Bashgal from their great village of Kombrom, which the maps call Kamdesh, and parts of Kunar; Strand derives their name from the ancient Kambojas, a people of these mountains named in Sanskrit and Persian sources, which is his hypothesis and he presents it as one. The Mumo live in three villages at Mandagal, about fifteen hundred people. The Kshto were driven from their village of Kushtoz by the Kom and now live scattered in the Dungal valley and elsewhere in Kunar. The Jashi, believed to be the valley's original inhabitants who once had a language of their own, live at Gawardesh and in Badugal, and the Binio and Jamcho are small groups attached to the Kom. Below all of them were the bari, the craftsmen who were once enslaved and remain a separate group.

The Kom and the Kata speak dialects of one language but do not claim a common origin. The Kom tradition, which Strand recorded from Qazi Ghulamullah and Vakil Muhammad Kabir in 1992, runs from Khorasan through Kandahar, Kabul and Kapisa to Kama at the meeting of the Kunar and the Kabul rivers, from which they were expelled about a thousand years ago, and then up the Pech to Kamatol before a massacre drove the survivors to the Bashgal. Their genealogy of named descent groups reaches back some twenty-nine generations. In January 1896 Amir Abdur Rahman's army under Ghulam Haidar Charkhi came up the Kunar, took the Kom elders hostage at Barikot, sacked Ushtrot, Pitigal and Saret and captured Kombrom, and the Kshto, Mumo and Kata were conquered after it. The sons of the leaders were taken to Kabul. Kamdesh was where George Scott Robertson had spent the year before, writing the fullest account of the old life, and it has its own pages on this site, as do Kantiwa, Bargi Matal and the 2009 battle below the village.