Tuesday, March 6, 2012

The Proto-Indo-European Homeland Puzzle

The Indo-European Language Family

Indo-European was the first language family to be identified. This discovery, and the beginning of modern linguistics, can be dated to February 2, 1786 at a gathering of scientists and other interested men. Sir William Jones, speaking at the Asiatic Society in Calcutta, made this astounding statement:

The Sanskrit language, whatever be its antiquity, is of a wonderful structure: more perfect than the Greek, more copious than the Latin, and more exquisitely refined than either; yet bearing to both of them a stronger affinity, both in the roots of verbs and in the forms of grammar, than could possibly have been produced by accident; so strong indeed, that no philologer could examine them all three, without believing them to have sprung from some common source, which, perhaps, no longer exists.

Jones later added Persian and Celtic as likely members of this family of languages.

Jones was uniquely qualified to make this discovery. His parental language was Welsh; he was taught English at school; he learned classical Greek and Latin in university where he studied law; he wrote the first English grammar of the Persian language (which earned him a reputation as one of the most respected linguists in Europe); and when appointed a judge in India at age 37 set out to learn the Sanskrit language to better understand local laws. Thus by age 40 Jones was familiar with a language in 6 (out of a total of 12) different Indo-European language branches.

Indo-European languages are spoken today by over 3 billion people - about half of the world's population - as either a first or second language. These languages are divided into 10 or 12 language branches or subfamilies. See the attached graph (Figure 1.1 of The Horse, The Wheel and Language p.12) which is arranged more or less geographically. English is a member of the Germanic subfamily along with German, Dutch, Frisian, the Scandinavian languages (which includes Icelandic), Yiddish, and Afrikaans. Other languages to note include:
            Tocharian – two extinct languages found in western China, the farthest East branch
            Hittite – a member of the extinct Anatolian branch – the earliest branch to separate
            Romany – the language of the Gypsies of Europe, is a member of the Indic branch showing that they originated in northwest India (not to be confused with Rumanian which is a member of the Latin or Romance language branch)

Source: Figure 1.1 of The Horse, The Wheel and Language p.12

About 6,000 to 5,000 years ago the parent language, called Proto-Indo-European, was spoken by a semi-nomadic tribe of people in the southern Ukraine and Russia. How their language spread and evolved into all of all these languages could be the subject for a future lecture. Today I want to show how historical linguistics and archaeology were combined to solve the puzzle of who the speakers of Proto-Indo-European were, and where and when they lived.
Source: Figure 1.2 of The Horse, The Wheel and Language p.14

The Proto-Indo-European Homeland Puzzle

Since the discovery of the IE language family, the location of the homeland of the original speakers has been claimed by different people to be many different places: India, Pakistan, Syria/Lebanon, the Caucasus Mountains, Turkey, Armenia, Kazakhstan, Russia, Ukraine, the Balkans and Germany. By the late 20th century linguists only seriously considered two of these – Anatolia (modern Turkey) and the steppes of southern Ukraine and Russia. And as recently as 2000, Calvert Watkins in his essay “Indo-European and the Indo-Europeans” which introduces his book The American Heritage Dictionary of Indo-European Roots stated “Archaeologists have not in fact succeeded in locating the Indo-Europeans.”

Colin Renfrew was a strong supporter of the other serious contender, Anatolia. Renfew's elegant proposal, published in the 1990's, had Proto-Indo-European migrant farmers carry their language along with agriculture from the Middle East to the westernmost part of Europe. But like many elegant theories, this one turned out to be not true. (I was greatly disappointed when linguistics and DNA analysis disproved Thor Heyerdahl's theories of Polynesian origins). There are, as we will see, serious problems with Renfrew's theory.

Before going further, I need to emphasize one point. Proto-Indo-European is a language. It is not a culture, nor is it a genetically-definable population. Language does not necessarily follow cultural boundaries, which can be determined by archaeology. Every first year archaeology student is taught “pots are not people”. But we know that someone must have spoken this language, and they must have lived in a particular place during a particular time. So while looking for the speakers of Proto-Indo-European we need to be careful of this constraint.
Clues from the Language

Since Proto-Indo-European is a language, let's look first at clues to the homeland from the language itself. The American Heritage Dictionary of Indo-European Roots published in 2000 contains 1350 reconstructed root words and several thousand more words based on these roots. These words have been painstakingly reconstructed by comparing similar words (called cognates) from the daughter languages over the more than 200 years since Jones' discovery. What can we learn about the people who spoke this language from their vocabulary?

they knew four seasons with snow in winter
they were not familiar with tropical plants or animals
animals include: wolf, lynx, elk beaver, otter, mouse, fish
birds include: crane, goose, duck, eagle, woodpecker
insects: wasp, hornet, fly, louse, bee, honey (mead)
domestic animals include: dog, cattle, sheep and horse
horses play an important role in the culture
they practiced spinning and weaving of wool
they knew metallurgy - copper
they knew of the wheel and used wagons or carts (weak link in Anatolian)
they knew of boats and oars - words like nav (navigate, navy) and rowing.
gift exchange is an important part of their culture
the guest-host relation was important –  *ghosti is the root of both host and guest (ghost originally meant visitor or guest)
they borrowed words from Proto-Uralic, another Eurasian language family, suggesting that the Proto-Indo-European speakers must have lived close to, and likely traded with, people who spoke Proto-Uralic who then, as now, live in northern Europe and Siberia (Hungarian is a member of this family found in Europe because of recent migration (~900CE).

The seasons and animals indicate a northern location either in or adjacent to a forest. The words for bee and honey place the homeland west of the Ural Mountains as honeybees do not occur east of there.

Clues to Dating Proto-Indo-European

Language can also help place the Proto-Indo-European speakers in time as well as location.

Agriculture was introduced to Europe between 6700 and 6500 BC while the wheel was not known until 3400 BC and woolen textiles sometime after 4000 BC. For the daughter language families to have similar words for the wheel and wool, they must have separated from Proto-Indo-European after their arrival. This effectively eliminates the Anatolian farmer immigrant theory. Besides, the two or three Anatolian languages were very similar to each other and spoken by only a small number of people in this area, which strongly suggests they are spoken by Indo-European speaking migrants to Anatolia, not by the ancestors of the language.

The domestication of the horse provides additional clues. Horses were hunted for meat by the people of the steppe for millennia before they were domesticated. They were first domesticated sometime after 4800 BC, a thousand years after cattle were introduced to the area. But they were raised for their meat only. During a cool dry period (4200-3800 BC) horses would have an advantage over cattle because they can forage for themselves during the winter. [Pioneer farmers in Saskatchewan like my grandfather often turned their horses loose for the winter to manage for themselves, rounding them up in the spring]. Riding of horses began on the steppes sometime before 3700 BC and had spread to Northern Kazakhstan, the Caucasus Mountains, and into Europe, by 3000 BC.

An important tool used in the dating of horse riding is bit wear on horse molars. The identification of tooth wear caused by bits of metal, bone, rope and rawhide, was pioneered by the author of The Horse, The Wheel and Language – David W. Anthony, and his wife, fellow archaeologist Dorcas Brown. There is an interesting Saskatchewan connection here. One of the experts they contacted was Hilary Clayton who began studying the mechanics of bits in horses’ mouths while working in Philadelphia, and then took a job at the Western Veterinarian College in Saskatoon. Anthony and Brown followed her to Saskatchewan in 1985 and viewed the X-ray videos she had made of horses chewing their bits.

Riding horses provided a significant benefit to herders in the steppes. A man on horseback could manage a herd of cattle or sheep much larger than a man on foot. With the much later advent of wheeled carts, about 3300 BC, the herders could carry with them tents, food and water allowing them to take advantage of the vast areas between the river valleys. This opened up the steppe much as the horse did to the plains of North America 5,000 years later.

Dating the Daughters

Language provides clues to timing in another way. Linguists can date, with more or less certainty, when each of the daughter language branches separated from the mother language. Here is a list of the branches, in the order of separation, with the approximate date (all BC) of separation (from Figure 3.2 The Horse, The Wheel and Language p. 57).

            Anatolian        4200
            Tocharian       3700 - 3300
            Germanic        3300
            Celtic / Italic   3000
            Greek / Armenian 2500
            Balto-Slavic    2500
            Indo-Iranian    2500-2200

Clues from Archaeology – The Kurgan Cultures

With the time line narrowed to the period 4000 to 2000 BC, it's time to look at the archaeological record and see who was living in the likely homelands and how well they fit with the linguistic clues. The archaeology of the Pontic-Caspian steppes was mostly carried out by Soviet scientists and published in Russian. These were not translated into English until the 1990s. Anthony was one of the first western archaeologists to study this work and relate it to the Proto-Indo-European homeland question.

Anthony found a close fit with the western steppe peoples who built huge burial mounds called kurgans. Their culture varied somewhat over the Proto-Indo-European time line and also geographically from place to place within this large area, but their overall cultures were similar, especially compared to the foragers to the north and east and to the sophisticated farming cultures to the west and south. They were semi-nomadic, raising cattle and sheep. Horses were important both for meat and for riding to manage their growing herds. They used wheeled carts. They mined their own ore and made their own tools and weapons of copper, tin and bronze.

Even more compelling is the evidence, from archaeology, of known migrations out of the steppes in the right directions and at the right times to account for the birth of the daughter language families.

1) to the west 4200-3900 (Anatolian)
2) to the east 3700-3300 (Tocharian)
3) to the west - several waves (Germanic, Celtic, Italic)
4) to north (Baltic, Slavic)
5) to the east and south (Iranian, Indic)

I should explain that by migration I do not mean large scale movement of people displacing existing populations along with their culture and language. This may have been the case with the Pre-Tocharians who made a remarkably long migration in one jump to the Altai Mountains 2000 km to the east (equivalent to the journey made by my grandparents from southern Ontario to Saskatchewan, but without the advantage of trains). Most if not all the other migrations were by small groups who, through some combination of trade or intimidation, became rulers of existing populations. They brought with them enough of their culture to be recognized archaeologically; and they brought their language which, for a variety of reasons, was adopted by the others and continued to spread long after they were gone.
Puzzle Solved

While there may be a few objections to his theory not yet satisfactorily answered, Anthony is convinced that the Proto-Indo-European Homeland puzzle has been solved.

Source: Figure 5.1 of The Horse, The Wheel and Language p.84

I want to finish with a quote from The Horse, The Wheel and Language  p. 464

Understanding the people who lived before us is difficult, particularly the people who lived in the prehistoric tribal past. Archaeology throws a bright light on some aspects of their lives but leaves much in the dark. Historical linguistics can illuminate a few of those dark corners.