The Gypsy Chronicles

Are Europe’s Roma descendants of India’s scheduled castes and tribes? Or of Mahmud of Ghazni’s prisoners-of-war?
Ancestry
A 1930s photo of Swiss nomads with their gypsy wagon
India’s Banjaras live nomadic lives just like Europe’s Roma
OF NO OBVIOUS ORIGIN An Andalusian gypsy dancer in traditional dress

The first scientific connection between European gypsies and India is believed to have been made 350 years ago. In the early 1760s, a Hungarian theology student at University of Leiden named Vályi István acquainted himself with three ‘Malabari’ students there. István was from a landowning family that employed gypsies, and he recognised similarities in the language spoken by both groups. He compiled a list of 1,000 such words from the Malabaris and took it back to the gypsies on his return from the university. They explained the meanings without difficulty. This is not the account of István himself. A printer soon came into possession of István’s notes and gave them to an army captain, who passed them on to an academic named George Pray. He found the discovery interesting enough to publish a note on it in a journal in 1776. The story eventually became lore among linguists. Except that István might not have existed.

In 1990, Ian Hancock, a professor at University of Texas and one of the world’s foremost scholars of Roma (the proper term for gypsies) studies, decided to find out more about Vályi István. Of Romani birth himself, Hancock has also set up The Romani Archives And Documentation Centre, which has the world’s largest collection of research material on the community; as an activist-academic, he has also been correcting popular misrepresentations of the Roma. He went to University of Leiden and found that its records of the period had no student with that name. Nor had anyone traced the original 1,000-word list. Also, Hancock notes in his essay on István: ‘Malabari, the language of the Malabar Coast, (at the time) referred to Tamil rather than to Malayalam, as it does now… But like Malayalam, Tamil is a Dravidian language and therefore quite unconnected with Romani…’

And so this was a double puzzle—not only was the central character missing, his story had fundamental errors. On the other hand, Hancock found that there was a Stephanus Waali at the University of Utrech about ten years earlier, in 1753, and that the Malabari students—probably from Dutch Sri Lanka, according to him—were all enrolled in Leiden during the same period. To complicate matters further, in the 1760s, there was a Hungarian theology student whose surname was Szatmari.

Hancock’s inference is that István probably visited Leiden from time to time, ‘perhaps to meet with other divinity students. It is also possible of course that it was not Vályi at all who should be credited with this discovery, but one of the Szatmaris, who were also Hungarian, and who may also have known some of the Romani language.’

There is no definite answer, but since that mysterious beginning, the corpus of research on the Roma’s Indian connections has kept growing. Linguists and historians have established beyond doubt that the Roma had their origin in India. To illustrate the linguistic legacy—if knife is churi in Hindi, it is churi in Romani; nose is nakh in Hindi and again nakh in Romani; and the numeral 20 is biš’ in Romani, just as it is bees in Hindi. Similar relationships can be found with other Indian languages. Or take the phenomenon of some Roma groups calling themselves ‘Sinti’, derived from the Sindhu or Indus River, just like Sindhi.

+++

In recent years, genetic studies have been done to reinforce what linguists have known. A few months ago, the findings of one that involved the Centre for Cellular and Molecular Biology (CCMB), Hyderabad, was published in the journal PLOS One. The objective of this study was to find ancestral sources of the Roma in India. The CCMB’s huge collection of over 20,000 DNA samples was an invaluable resource. When Gyaneshwar Chaubey, a genetic researcher with University of Tartu in Estonia, came to Hyderabad in 2009, he had discussed doing this study with CCMB scientist Kumarasamy Thangaraj.

The CCMB has been conducting genetic studies to gain an understanding of India’s population history. Because of the country’s caste system, which keeps DNA locked within communities, India is fertile ground for such research. The CCMB has identified more than 4,600 Indian ‘populations’—those who marry within groups—and has been working on a large number of them.

Genetic studies of populations are done by studying mutations in DNA specific to a group. Y chromosomes are passed along only by men and so a person’s paternal ancestry can be traced by studying mutations in it. Similarly, mitochondrial DNA (mtDNA) is transmitted through women and studying it can throw light on maternal ancestry. Depending on their DNA mutations, individuals can be categorised into sets called ‘haplogroups’. Regions, castes, tribes, communities and so on can have haplogroups associated with it. ‘Haplogroup H1a1a originated in India and is now carried by around 55 million people of this country. Our task was to find a group of people closest to the European Roma carrying this mutation,’ wrote Chaubey in an email interview.

“We analysed a large number of [samples from the] Indian population, particularly those who carry the H1a1a haplogroup. We used additional markers to see which groups are closer to the European Roma,” says Thangaraj. For the study, they first separated the samples with H1a1a mutation. Then they picked 5 to 10 individuals from different population groups and looked for Standard Tandem Repeat (STR) markers. Human DNA codes are made up of four nucleotide bases—adenine (A), guanine (G), cytosine (C) and thymine (T). STRs are repeats of these bases in DNA. For example, AGCCAGCCAGCCAGCC is a repeat of the AGCC base. The number of repeats can vary between individuals. ‘We used individual-wise repeats and compared them [with Romani samples]—the idea was that individuals sharing close ancestry would have less difference,’ says Chaubey.

What they found was that the scheduled caste (SC) and scheduled tribe (ST) populations of northwestern India had the least differences of these repeats with Roma individuals. “Groups in northwestern India, particularly SC and ST ones, showed much closer genetic affinity with Europe’s Roma,” says Thangaraj.

This finding is not novel for linguists and historians in the field. One of the basic conflations has been that the word ‘Roma’ itself comes from ‘Doma’, an umbrella term for some SCs and STs in India. Like SC/STs, the Roma too have been victims of discrimination in various societies they passed through; gypsies were among the groups that Hitler tried to annihilate. Despite all this, the question of common ancestry is far from settled.

+++

There are two theories about why the Roma first left India. In the traditional view, they were nomads related to India’s Doms in some way, and in the course of their itinerancy, moved from central India to northwest Punjab and Kashmir. From there, they went to the Middle East but probably did not spend a lot of time there because their language did not have much Arabic influence. Then they moved to Turkey (or the Byzantine Empire), and from there to Europe. There is no unanimity on why or how or even when all this began. A 1988 paper by a Hungarian linguist, Jozsef Vekerdi, describes gypsies as ‘leaving northwest India probably about the seventh century AD. They are characterized as robbers, murderers, hangmen and entertainers…they belonged to the so-called ‘wandering criminal tribes’ of India and were obliged to lead a parasitic way of life. Among the numerous outcast groups, they occupied the lowest rung on the social scale.’ Another 1997 paper says, ‘All groups considered Gypsies are descended from ancestors who left India around 900 AD….’ A 2005 paper by an American scholar, Kenneth Blachut, traces the exodus to the 6th century AD. He writes: ‘Gypsies had begun leaving the southern part of India in 1500 BC, when the Aryans invaded the country . . . Gypsies originated in the Punjab region of northwestern India. They began leaving in the sixth century AD because of constant invasions by Islamic warriors.’

The other hypothesis of the exodus, one that is gaining currency, is propounded by Hancock. He attributes the mass migration to Mahmud of Ghazni’s invasions of India in the 11th century AD. The Hindu rulers at the time raised an army to resist these invasions. After their defeat, the soldiers and camp followers who were taken as prisoners-of-war eventually became the Roma. ‘Only a few of those people were actual kshattriya warriors; most of them were service providers, camp followers,’ says Hancock in an email interview.

The Ghaznavids were also at war with the Seljuk Turks for control of the empire. In 1038 AD, the Turks defeated them and liberated the Indians. Both together then defeated the Kingdom of Armenia 30 years later. It was there in Turkey that the Indian groups melded to become the Roma community. Earlier in their camps they had evolved a rough language to communicate among themselves. This language became Romani, a mother tongue. When the Ottoman Empire of the Turks expanded into Europe, the Roma also followed.

According to Hancock, the PLOS One study that genetically links the Roma with SC/STs is just another hypothesis. ‘It doesn’t take social and historical factors into consideration,’ he says. ‘Several other genetic studies on my people have been done which give quite different results; Bhalla, for instance, found that the Roma descend from Jats; Kochanowski finds a Rajput origin.’

An interesting element in support of the Ghaznavid postulate is a study of India’s Banjara community. Vijay John, a student of Hancock at University of Texas, is currently working on a dissertation on the origins of the Romani people by looking at the linguistic, socio-historical and genetic evidence. “Banjaras have an oral legend that their ancestors were taken by Mahmud of Ghazni as prisoners-of-war. The Banjara version is that they are descendants of the soldiers who stayed behind. Of those who were not taken,” he says.

Linguists don’t just rely on such narratives. They also have more sophisticated techniques. For example, to pin down the time period of the exodus, one of the things that Hancock studied was grammatical gender in the languages (the treatment of nouns as masculine or feminine; English does not have it, but many languages do; in Hindi, for example, ‘vehicle’ is feminine and so you say ‘gaadi chali gayi ’). Sanskrit has three genders—masculine, feminine and neuter. But as the Indo-Aryan languages evolved from it, the neuter gender was lost (around 1000 AD). In languages like Hindi and Punjabi, the neuter merged with the other two genders. The word for water was originally neuter in Sanskrit. But its Hindi descendant pani is masculine. The Romani descendant too is pani, also masculine. The word for eye is aksi in Sanskrit, which was originally a masculine noun. In Hindi, its descendant ‘ankh’ is feminine. In Romani, the word for eye is yakh, which is also feminine.

That is an indication of a shared switch in gender references. The argument goes like this: if the same change occurred in both Romani and the Indian language, then it must have happened before the Roma left India. “Which would mean that they couldn’t have left earlier than 1000 AD (since that was when the neuter was lost),” says Vijay. He is also conducting research on the origin of Romani numerals (of which 1--6, 10, 20 and 100 are clearly of Indian origin). Vijay is looking at which languages have a form closer to Romani for a given numeral. And he then plots on a map where those regions of India are. This is then used to make deductions about the origins of the Romani people and language.

+++

The political aspects of Roma studies usually relate to the discrimination they have suffered, and it so happens that theories about origins are often used as tools by racists. The root causes of European discrimination against the Roma are many, starting with their arrival in Europe that coincided with Islamic expansion. They were therefore associated with a threat to the Christian world even though they were neither Christian nor Muslim. They were dark-skinned and medieval Europeans associated darkness with sin. Since they were not bound to any land, the concept of nationalism was alien to them. Fortune-telling, an easy source of income for nomads, was respectable in India but viewed with suspicion and fear in Europe. The Roma also kept a distance from others. ‘We have retained aspects of the caste system and regard non-Roma as spiritually polluting,’ says Hancock.

Discrimination against the Roma is still widely prevalent. In a 2003 poll in the UK, 35 per cent of respondents admitted a prejudice against gypsies. Linking the Roma with India’s SCs and STs provides, as Hancock puts it, ‘White racists in Europe with a lot of fuel!’