Wanetsi (Tarino): the speech of the Spin Tarin of Harnai

At a glance

• Wanetsi (Waṇetsī, وڼېڅي), which its speakers call Tarino, is spoken by sections of the Spin Tarin around Harnai, Spin Tangi and Sanjawi in north-eastern Balochistan. [1, 2, 5]

• Georg Morgenstierne, who recorded it in 1929, found it stood entirely apart from all other known varieties of Pashto. Josef Elfenbein wrote that a fair case could be made for calling it a separate language. [1, 3]

• Its speakers call it a dialect of Pashto. Ibrahim Khan, a barrister writing in the Pashto Academy Peshawar's journal, says so, and so does the cover of Umar Gul Askar's book on it, published by the Pashto Academy in Quetta. [2, 3]

• No one has counted its speakers. Estimates run from "a few thousand or less" (Elfenbein, 1972) to 25,000 speakers (a speaker told Morgenstierne) and 30,000 homes (a speaker interviewed for a survey published in 1992). [3]

• Elfenbein, Morgenstierne and Skjærvø place its split from the rest of Pashto in the Middle Iranian period, long before the written language. [3, 5]

Where it is spoken

Wanetsi is spoken in a small mountain territory east of Quetta. Josef Elfenbein, writing in 1972, described it as about 20 miles long and 15 broad, oblong in shape and, except for the Harnai valley, lying entirely in high mountains. [3] Mountains split it into a northern and a southern part, and between them the only interior road runs from Harnai to Samalan. [3]

The southern part runs along the railway line from Shahrig, about 40 miles east of Quetta, for some 20 miles up to Spin Tangi. Shahrig itself now speaks ordinary Pashto. In 1929, Elfenbein wrote, it still lay inside Wanetsi territory, and the language now begins about five miles east of it. [3] The northern part lies along the Quetta to Loralai road beyond Ziarat. Wanetsi villages start about six miles east of Ziarat and the last are in Sanjawi tehsil, about 20 miles south-west of Loralai. [3]

The Pashto Academy's own account agrees with this map. The cover of Umar Gul Askar's book on the language calls it a "remote Pushto dialect spoken in Harnai Tehsil of District Sibi and Sinjawi Tehsil of District Loralai". [3] Ibrahim Khan, drawing on a Pashto-language book on the Tarin by Arif (2011), says the Wanetsi clan lives mainly in and around Chawter in Sanjawi and the Makhiani clan in Harnai, with territory from Khost (the Khost of Harnai district, not the Afghan province) to Spin Tangi and as far as Tsandi. [2]

Denys Bray, the British officer who wrote the 1911 census report for Balochistan, put the Wanetsi and Makhiani at "Chairing and Duki", in the words quoted by Hallberg. [3] Elfenbein, Khan and Askar all place the language around Harnai and Sanjawi.

Names and people

The language has three names. Elfenbein found that in the north it is universally called Tarino, though some speakers preferred Wanetsi, after the Tanda Wani district. Speakers living in Quetta, who often call themselves Makhiani, also call it Chalgari. [3] Khan uses Tarino, because that is the name its speakers use for it. Linguists know it as Wanetsi. [2]

The speakers are mainly Spin (White) Tarin, one of the two divisions of the Tarin along with the Tor (Black) Tarin, and in part people of other tribes. [1, 3] Khan names three branches that speak it, on the word of the poet Nizamuddin Nizami of Chawter: the Wanetsi, the Makhiani, and the Petsi Shekh, who are settlers in the other two's territory. [2] He also corrects a colonial reference book. The General Staff's 1910 Dictionary of the Pathan Tribes calls the Makhiani Tor Tarin, and Khan, following Arif, says they are Spin Tarin. [2]

How outsiders first recorded it

Before 1929 almost nothing was written about it. The Gazetteer of the Sibi District and Bray's census mentioned it and quoted a handful of words. [1] Morgenstierne made a point of finding a speaker in Quetta and stopped for a day in Harnai, and he published a first description in 1930. [1, 5] Elfenbein followed with a long study in the journal Archiv Orientální in 1967, a description in the journal Afghanistan in 1972 and a two-part paper, "The Wanetsi Connexion", in the Royal Asiatic Society's journal in 1984. [2, 3, 5]

Bray, writing as a British census officer, was dismissive. Hallberg quotes him from Grierson's Linguistic Survey: he called it less a Pashto dialect than a corruption of one and argued that the Spin and Tor Tarin were Indians, possibly Jat, who had joined the Pathans without yet taking up their language. [3] Later linguists rejected that reading, and Hallberg calls the view of Morgenstierne, Elfenbein and Skjærvø "quite different". [3] Skjærvø put it this way: the Pashto area split into two dialect groups before there was any written Pashto, one represented by all the dialects of modern Pashto and the other by Wanetsi. [3]

Sounds

Most of what makes Wanetsi strange to a Pashto ear is in a few regular changes that Morgenstierne listed. [5]

• A t between vowels is lost or becomes y. Standard Pashto turns the same old consonant into l. So Wanetsi piyār, "father", stands against Pashto plār, and sī, "hundred", against Pashto sal. [1, 5]

• The old group ft becomes w: tōw (feminine tawa), "hot", where Pashto says tod. [5]

• An r before ž is kept: yirž, "bear", and tərža, "thirsty". [1, 5]

• Wanetsi does not palatalise: māst, "month", and atā, "eighty". [5]

• The word for "twenty" is šwī. Morgenstierne derived it and Pashto (w)šəl from the same old form but argued that the Wanetsi word cannot come from the Pashto one. [1, 5]

• The vowel shift that marks the Karlanri tribes of Waziristan and Bannu (ā, o and u moving to o, e and i) is absent. Khan compares the Bannu word for "father", plor, with Wanetsi piyār and lists Wanetsi among the south-eastern dialects the shift has not reached. [2]

• A final -ay on masculine nouns, kept in Kandahar speech, becomes a stressed -á. Khan gives saṛá for "man" and woṛá for "summer". [2]

Anna Boyle David's table of Elfenbein's 1997 scheme puts the Tarin and Wanetsi in the south-eastern group with the Sherani, Bannu, Waziri, Kakari and Achakzai. [4]

Grammar

Wanetsi marks possession with a particle placed after the noun, γa, where Pashto puts da before it. [1] Morgenstierne counted the da form as rare in Wanetsi and probably borrowed. [5] The pronoun "we" is mōš, where Pashto has mūnğ, and he took the lack of the Pashto nasal as a sign of early separation. [5]

The verb has its own endings. Morgenstierne gave kī, "I make", and a separate third-person plural kīn, and noted how few endings there are: all personal forms except the third plural are carried by whatever final vowel the word has. [5] The present often takes an infix, -en-, as in murš'enī, "I rub". [5]

Khan, writing in 2021 from recordings of Nizami, gave the first-person ending as ī, set against the nasal ə of the Karlanri speech of Bannu and Waziristan. His example, with the Pashto script he gave:

زې اوبه غوزي

ze obə ğozī

"I am drinking water." [2]

Two more of his examples show verb stems that Wanetsi shares with its neighbours. The stem drəm- is a form of "go", as in:

ته اروته مه درمه

tə órta mə drəma

"You don't go over there." [2]

The imperfect past of "go" in Wanetsi is īd-, as in the Orakzai speech:

زه ایدي

zə īdī

"I was going." [2]

Some words

Morgenstierne listed words that Pashto does not have: γoz-, "to drink", mīt, "fist", wūn, "naked", zūng, "knee", and the feminine kəžə, "big fish". [5] Khan's Wanetsi speaker gave orboy for a courtyard or haveli and woṛá for summer. [2]

Morgenstierne's sample sentence runs:

Indī waguṛī čī mōš piyār γa caṭ lēždī wī

It means "in this village our father had many bulls". [5] Waguṛī is "village" and lēždī "bulls". [5]

A second, in the same article: šamze o xwāržə šwe mī de γōzīn, "they also drink buttermilk and sweet milk". [5]

Dialect or language

Morgenstierne wrote in 1932 that Wanetsi had some claim to be considered, in its base, a separate Iranian language, though closely related to Pashto and covered with Pashto elements to such an extent that it is often hard or impossible to pick out the genuine Wanetsi forms. [1] In the Encyclopaedia Iranica in 1983 he called Pashto essentially one language, with Wanetsi the one possible exception. [5] Elfenbein wrote that it was so different from every other Pashto dialect that a fair case could be made out for a separate language. [3]

The wordlist evidence is closer than that suggests. In a survey published in 1992, Daniel Hallberg compared a Wanetsi list from two men from Harnai, living in Quetta, with Pashto lists from thirty-four places. The Wanetsi list was 71 to 75 per cent similar to those of Balochistan and Kandahar, and 63 to 72 per cent to the rest. [3] That was about four points further from the other Pashto varieties than Waziri, the most distant of them, was. [3] Ten of the twenty-six Pashto speakers Hallberg interviewed in Balochistan named Harnai as a place where Pashto is spoken very differently. [3] The president of the Pashto Academy in Quetta told him it was hard to understand and did not seem sure whether to call it a language or a dialect. [3]

The speakers see it differently from the linguists. Khan notes Elfenbein's remark and adds that the speakers identify their speech as a dialect of Pashto, not as a separate language. [2] Hallberg's interviewee, a man from the village of Tuka east of Harnai, preferred the name Tarino. Asked what was good about it, he said: "It is sweet to us." [3]

Links with the Karlanri tongues

Khan's article asks why the language of the Spin Tarin shares so much with the Karlanri dialects of Bannu, Waziristan and Khost, which lie far to the north-east. He does not claim an answer and leaves it to the reader. [2] He lists what they share. The word for "rain" is wūr in the Tarino he recorded and wör in Waziri, a word he says other dialects lack. [2] His speaker found about fifty words in common on a visit to Bannu. [2] The verb stems drəm-, "go", and kəs-, "look", appear in Karlanri speech and in Tarino. [2] Khan says in the article that he is not a linguist, and that the thought that wūr is the older Pashto word for "rain" is a guess until an expert agrees. [2]

A language under pressure

Morgenstierne said in 1932 that Wanetsi had been pushed back and made more Afghan, and that a few Wanetsi words were used by Kakars whose homes were well north of Harnai, which showed the language had once covered more ground. [1] In 1983 he wrote that it was "being more and more influenced by and pushed back by ordinary Paṣ̌tō". [5] Elfenbein found the territory smaller than in 1929: Shahrig had gone over to Pashto. [3]

The speakers' own account in Hallberg's survey was confident. The Tuka man told Hallberg that no one had stopped speaking it, that his children would use it most and that it would not die out. He added that the people of Harnai could not speak Pashto and that only town people could. [3] Hallberg wrote that more study was needed, in particular of how many active speakers there were, how Tarino speakers use Pashto and Urdu, and how well they speak them. [3] The Voice of America's Pashto service headed a report on the Tareeno dialect of the Pashtuns in Balochistan with the words that scholars say it is dying out. [6]