Why Do Some Pashtuns Have Light Eyes?

We are Nuristanis, and the question asked about Pashtuns is the one asked about us. Why are some of them fair. Where did the blue eyes come from. Were there Greeks. We have written about the Afghanistan-wide version of that question already, and this page is about the Pashtuns in particular, because they are the largest people in the country, our neighbours on three sides, and the ones the internet asks about most.

There is a reason to be careful with it. The same question has been asked for two hundred years by people who wanted a particular answer, and the answers they produced did real harm. So this page sticks to what has been measured, says who measured it and when, and says plainly where the evidence stops.

The short answer

• Most Pashtuns have brown eyes and dark hair. In the one study that counted, about 82 per cent had brown eyes.
• About 5.6 per cent had blue eyes and 12.8 per cent green or hazel, so the in-between shades are more than twice as common as clear blue.
• A Soviet survey of nineteen Pashtun samples in the 1960s found the same pattern by a completely different method: truly light eyes close to nil everywhere, and what varied between groups was the share of mixed eyes.
• The variant behind blue eyes is a single old change in one regulatory switch, shared by every blue-eyed person on earth. It is present at low frequency right across this region.
• Pashtun skin is on average slightly lighter than that of Baloch and Punjabi neighbours, measured with an instrument rather than by eye.
• No study has shown Greek descent beyond two individual men, and the famous study said to prove Israelite descent was never published.

What has actually been counted

Only one study has ever photographed and classified Pashtun eyes. In 2023 a team led by Muhammad Adil Rahat published work on 893 people in the Malakand division of northern Pakistan, in the districts of Swat, Buner, Shangla, Malakand, Upper Dir, Lower Dir and Chitral. Each person's eyes were photographed in a light box and sorted by trained observers into brown, intermediate and blue. The result was 81.6 per cent brown, 12.8 per cent intermediate and 5.6 per cent blue.

Two honest caveats go with those numbers. The sample is described as the local Pakhtoon population, but the paper records no per-person check of ethnicity and notes that Kohistanis and Gujjars also live in the area, so it is a Pashtun-majority sample rather than a verified Pashtun one. And no study of this kind has ever been done in Afghanistan, where most Pashtuns live. There is no Afghan figure to quote.

The same study tried to predict those eye colours from DNA, using the system forensic laboratories use, and the result is the most useful thing in it. Blue and brown were predicted correctly almost every time. Green and hazel were a coin toss: the system kept calling them blue or brown. That is because the prediction systems were built on European populations, where the in-between shades are less common. It is the scientific reason that talking about Pashtun eye colour in numbers is harder than it looks, and why the greenish middle of the range, which is where most light-eyed Pashtuns actually sit, is the least well understood part of it.

The survey nobody reads, and what it found

Long before DNA there was a survey, and it is the only large body of measurements from Afghanistan itself. Between 1964 and 1968 the Soviet anthropologist Georgy Debets and colleagues from Kabul University recorded the eye and hair colour of men across the country using printed colour standards. His tables include nineteen Pashtun samples, from Kunar, Nangarhar, Laghman, Wardak, Khost, Ghazni, Zabul, Uruzgan, Kandahar, Spin Boldak, Farah, Herat and among the Durrani and Ghilzai.

Across those nineteen samples the share with dark eyes runs from 55 per cent, among the Pashtuns of Wardak and the Ghilzai, to 86 per cent, among the Pashtuns of Kunar. That sounds like a lot of variation until you work through the means he also printed. In the lightest-eyed sample, where 55 per cent scored dark, the rest average out as almost entirely mixed, with the truly light category close to zero. The same holds for every other sample. What changes from one Pashtun group to another is not how many have blue eyes, it is how many have the greenish and hazel shades in between.

That is exactly what the Malakand study found with photographs and a light box fifty-three years later: 5.6 per cent clearly light, 12.8 per cent in between. Two completely different methods, two generations apart, on different sides of the Durand Line, agreeing about the shape of the thing. It is the firmest finding on this page.

What the geneticists have found

Blue eyes are, in effect, one variant in one genetic switch. A Danish team established in 2008 that a change in a regulatory region of a gene called HERC2, which turns down a neighbouring pigment gene, is almost perfectly associated with blue eyes, and that blue-eyed people in Denmark, Turkey and Jordan carried the same version of it. Wherever blue eyes turn up in the world, it is the same old variant, inherited in a near-recessive way, which is why two brown-eyed parents can have a blue-eyed child. A blue-eyed Pashtun and a blue-eyed Dane share one distant ancestral change and nothing else of consequence.

How common is that variant among Pashtuns? A Danish and Pakistani team measured it in 2021 in 103 Pashtuns from Khyber Pakhtunkhwa alongside 107 Baloch and 89 Punjabis. The blue-associated version was carried on about 16.5 per cent of Pashtun chromosomes, against about 6.7 per cent among the Punjabis and 6.5 among the Baloch. Other datasets put the Pashtun figure between roughly 16 and 22 per cent. For comparison, it sits above 60 per cent in Europe as a whole and near 77 per cent in north-western Europe, and near zero in East Asia. So Pashtuns carry it two to three times as often as their immediate neighbours and a quarter as often as Europeans. That is the entire size of the thing being explained.

The same study measured skin pigmentation with an instrument rather than by eye, on the upper arm, lower arm and forehead, and found the Pashtun group significantly lighter than the Baloch and Punjabi groups at every site. It is a real average difference, and it is an average: the ranges overlap heavily and no one person is described by it.

The steppe was not blue-eyed

Here is the finding that quietly disposes of most of the mythology. The popular picture has fair, blue-eyed horsemen riding down from the steppe into Iran and India in the Bronze Age and leaving their colouring behind. Ancient DNA says otherwise. A 2014 study recovered pigmentation data from skeletons of the Pontic-Caspian steppe, the very population in that picture, and found the blue-eye variant on only about 16 per cent of their chromosomes, against about 65 per cent among people living in the same region today. Blue eyes were commonest much earlier, among the hunter-gatherers of Europe, and rose in Europe afterwards, by selection, over the last five thousand years.

Pashtuns do carry a share of steppe ancestry, as most people of northern South Asia do. But if the steppe people themselves carried the blue-eye variant at about 16 per cent, then steppe ancestry predicts a modest elevation in Pashtuns over their southern neighbours, which is precisely what is observed, and nothing more dramatic. The arithmetic works out to the ordinary answer.

The rest of the explanation is the one that also explains Nuristan. In a population of millions a variant at one in six chromosomes shows up here and there. In a valley of a few thousand people who marry among themselves, chance alone can push a trait several times above or below the regional average, which is why light eyes cluster in particular families and villages. Marriage within the family or the clan, common across this whole region, does not create such variants; it makes a near-recessive one surface more often where it is already present. That is why the honest answer to why does that family have green eyes is usually about that family, not about their nation.

The Greeks

The story that the fair people of these mountains descend from Alexander's soldiers is attached to many peoples here: to the kings of Badakhshan, to the rulers of Hunza, Nagar, Skardu and Chitral, to the mountain Tajiks, to the Kalash, to the Nuristanis and to the Pashtuns. It is old, and in most cases it can be shown to be younger than Islam in the region, which places it a thousand years and more after Alexander.

For Pashtuns there is one real piece of evidence, and it is small. A 2007 study of 952 men, including 77 Greeks, found two Pashtun individuals carrying a lineage common in Greece; one of them shared a detailed sixteen-marker profile with three Greek men and the other differed by a single mutation. The authors called this a strong indication of a European, possibly Greek, origin for those particular chromosomes, while concluding in the same paper that a large Greek contribution to any Pakistani population is excluded. A larger survey of 190 Afghan Pashtun men in 2012 did not find that lineage at all. Two men is not a people. An earlier study using different markers found a small Greek-like signal among Pashtuns and Burusho that it judged probably older than Alexander's invasion.

What is better documented is how the idea arrived. The British mission to Kabul in 1809 went looking, in Mountstuart Elphinstone's own words, for our Macedonian neighbours, having read of a tribe in Bajaur that claimed descent from Alexander's army. They were told of a people in the mountains to the north who were celebrated for their beauty and their European complexion, drank wine from silver cups, used chairs and tables and spoke a language nobody around them understood. Elphinstone published that in 1815, and then demolished it himself: the languages of those mountains, he wrote, have a near connection with Sanskrit, and this derivation of their language seems fatal to the descent of the Kafirs from the Greeks.

He also corrected something still misread today. The two divisions of the Kafirs were called the black and the white not for their skin but for their clothes, one wearing black goatskin vests and the other white cotton.

The theory went on circulating anyway. Seventy-five years later Surgeon-Major Henry Bellew was identifying Greek descendants among the Pashtun tribes by the sound of their names, and writing of one group near Kabul that they were sometimes quite as fair as Englishmen; at least, he added, I have seen two such. That sentence is the whole evidential basis of a chapter. Bellew was also clear about why the work mattered to him: the study of these different peoples, he wrote in 1880, is of the first importance in view to their ere long becoming subjects of the British Empire.

The children of Israel

Many Pashtun families hold that their ancestors were of the Bani Israel, the children of Israel, carried into captivity and settling eastward through Persia into Ghor. Bellew relayed the tradition in 1880, and it is far older than he is, recorded in Pashtun genealogies for centuries. It is held with conviction, and it belongs to the people who hold it.

What can be said is that the genetic studies which have looked for it have not found it. And the study most often cited as proving it, announced to newspapers in 2010 and much repeated since, was never published; the names usually attached to it in later accounts belong to other researchers' work. A tradition of descent says something true about how a people understand themselves and where they place their history. It is a different kind of claim from a laboratory result, and the honest thing is not to pretend either one settles the other.

Sharbat Gula, who had green eyes

The most famous light-eyed face in Afghanistan belongs to a Pashtun woman, and almost everything popularly said about it is slightly wrong. Sharbat Gula was photographed by Steve McCurry in 1984 at the Nasir Bagh refugee camp near Peshawar, as a schoolgirl of about twelve, and the portrait ran on the cover of National Geographic in June 1985. She was born about 1972 in Nangarhar province; her parents were killed in a bombing when she was about six.

Her eyes are green, not blue. Every account that describes them, including the reporting of her later life, calls them green, and green is precisely the intermediate shade that the genetics handles worst and that the Malakand study found to be more than twice as common as blue. When a National Geographic team went to Afghanistan in 2002 to find her, several women claimed to be her, and she was identified by the pattern of her irises rather than their colour.

What happened afterwards is worth knowing too, because it is rarely attached to the image. In 2016 she was arrested in Peshawar for living in Pakistan on forged papers, convicted, fined 110,000 rupees, held for fifteen days and deported to Afghanistan at Torkham. Afghanistan is only my birthplace, she said then, but Pakistan was my homeland. In November 2021, after the Taliban took Kabul, she was evacuated to Italy at her own request and granted refugee status. It has also been alleged that the original photographs were taken without her permission. We have not reproduced the image on this page. A girl in a refugee camp became the most recognised Afghan face in the world and had very little say in any of it, and the least that can be done is to write about her as a person with a life rather than as a pair of eyes.

A word about the word Aryan

This subject cannot be discussed in English without the word turning up, so it is worth being exact. Aryan is properly a term for languages: the Iranian and Indo-Aryan languages, whose ancient speakers used the word of themselves, and to which Pashto belongs, as do Persian and Hindi. The Nuristani languages form a third branch of the same family. In the nineteenth century the word was lifted out of linguistics and reinterpreted as a racial category, and the United States Holocaust Memorial Museum's own account of the term says exactly that: a linguistic category reinterpreted incorrectly as referring to ethnicity or race, then adapted and radicalised into a belief in an Aryan race.

The professional bodies have been blunt about where that leads. The American Society of Human Genetics stated in 2018 that genetics demonstrates that humans cannot be divided into biologically distinct subcategories, and that it exposes the concept of racial purity as scientifically meaningless. The American Association of Biological Anthropologists stated in 2019 that race does not provide an accurate representation of human biological variation and that no group of people is, or ever has been, biologically homogeneous or pure. Someone with grey eyes in Swat and someone with brown eyes in Kandahar differ in a handful of variants out of many thousands, and in nothing else that matters.

What it looks like from here

Our own province sits at the top of the same valleys, and the same survey measured us. Debets found the Nuristanis of Kamdesh with 80 per cent dark eyes, which puts them inside the Pashtun range and darker-eyed than nine of his nineteen Pashtun samples, and he wrote that their pigmentation is not lighter than that of the Pashtun. He was arguing against the claim that European migrants stood among the ancestors of the Afghan people, and he preferred the explanation an earlier scientist had offered, that such peculiarities arose under the influence of isolation. A Soviet physical anthropologist, working inside the framework that produced these stories, using his own tables to take them apart.

There is one more thing Elphinstone saw in 1815 that is worth ending on. Among the eastern and the western Afghans alike, he wrote, men as swarthy as Indians, and others as fair as Europeans, are to be met with in the same neighbourhood. That has always been the reality here. Every valley holds every shade, and the interesting questions about Pashtuns, and about us, are about language, history, poetry, hospitality and how people have survived in these mountains. The colour of anyone's eyes is the least of it.