This is an extract from Our Human Story, our newsletter about the revolution in archaeology. Sign up to receive it in your inbox every month.
Every so often, I’m reminded of a truly startling fact. It is possible to travel from the UK, where I live, east through Europe into Ukraine and western Russia (maybe don’t do this bit), cross the Caspian Sea, then carry on east all the way to India – and always speak to people whose language belongs to the same family.
That’s because English is one of more than 400 Indo-European languages. French, German, Bengali (Bangla), Sanskrit, Albanian, Greek and many others are all Indo-European. In other words, all these languages are believed to descend from a single parent, dubbed Proto-Indo-European, that was spoken thousands of years ago.
Advertisement
But where was the Indo-European homeland, the place where Proto-Indo-European was spoken? And how did this one language family spread so widely?
There’s been a lot of research into this of late, so, to get myself up to speed, I read the new book by archaeologist and linguist J. P. Mallory. He has spent a large chunk of his career exploring the origins of Indo-European languages, and the book is a comprehensive account of our search for this mystery place. The quest, as Mallory explains, began in the late 1700s when the Indo-European language family was first recognised.
This isn’t a review, but I’ll tell you what you need to know. Are you interested enough in the homeland of the Indo-European languages to read 400-odd pages about that single question? If the answer is yes, this is the book to read: it’s comprehensive, up to date, clearly explained (except for some furiously technical bits around the two-thirds mark) and sometimes funny. That’s all you need to know. With that said, I am going to spoil the ending.
A lost proto-language
Linguists like William Jones (1746-1794) and James Parsons (1705-1770) discovered Indo-European by spotting similarities between radically different languages. Mallory lists some words that are alike in seemingly remote languages. For example, “eat” is á»ĺłľľ± in Sanskrit, Ă©»ĺĹŤ in Greek and ±đ»ĺĹŤ in Latin. Later, when archaeologists discovered the extinct Anatolian languages that were once spoken in what is now Turkey, those also turned out to have been Indo-European.
What we don’t have is any direct records of Proto-Indo-European. It existed in a time before writing. Linguists can reconstruct aspects of it by comparing the most distantly related Indo-European languages, along with the oldest written records, and working backwards – but you can probably imagine how tricksy this becomes.
The exact timing of the origin of the Indo-European family is interestingly uncertain. It began spreading at least 4500 years ago – it may well have been over 5000 years ago, and some estimates place it over 8000 years ago.
That is a nastily large uncertainty range. It’s perhaps no surprise that there has been a blizzard of suggestions.
Some can be safely rejected: Mallory mentions a gloriously cracked proposal from 1966 that the speakers of Proto-Indo-European lived at the South Pole, where “a great world civilization had existed until Antarctica was glaciated somewhere between 10,000 and 15,000 years ago”. (Note to readers: Antarctica became glaciated millions of years ago.)
However, there have been plenty of serious suggestions. Mallory mentions, among others, the Indian model, the Baltic model, the Nordic model, the steppe model and the Anatolian farmer model. As he says, “the Indo-European homeland has definitely been located”, but that’s because the range of ideas put forward “pretty well covers all the options”.
At this point, you may be thinking: well, surely the answer is obvious? Indo-European languages are spoken from the UK, France and Spain in the west, all the way to India and Bangladesh in the east. Surely the homeland was somewhere in the middle, and the languages spread out in all directions? That would put the origin point near the Black Sea and the Caspian Sea, perhaps modern Ukraine or Turkey or thereabouts.
This is an incredibly basic argument, so simple that an 8-year-old would probably think of it, and so it gives me great pleasure to tell you that it appears to be essentially correct. Journalist Andrea Valentino alighted on this general area in a żěè¶ĚĘÓƵ feature in 2022, because multiple lines of evidence – genetic as well as linguistic – pointed that way.
With the advent of methods to analyse ancient DNA, Mallory explains, the search for the Indo-European homeland has narrowed in on three competing scenarios that all place it somewhere in central Eurasia.
Lost Eden
The Anatolian farmer model focuses, as you might imagine, on early farmers in the region of Anatolia – what’s now Turkey. Many of them moved out into neighbouring areas, and this could have spread Indo-European languages over Europe and Asia. However, there are a few difficulties. The movements of people, as revealed by genetics, don’t seem to line up with the movements of languages. In particular, the Anatolian languages were the first to break away from the other Indo-European languages, which seems odd if the entire language family began in Anatolia.
Alternatively, Indo-European languages could have arisen in the Caucasus – the region between the Black Sea and the Caspian Sea, which today includes countries like Armenia and Azerbaijan. This would make sense, except there ought to be a trail of Indo-European languages throughout the Caucasus – and there aren’t. Almost all the languages spoken there aren’t Indo-European.
Finally, we get to the steppe model. To be specific, the Pontic-Caspian steppe to the north and east of the Black Sea, spanning much of modern Ukraine and its neighbours. Swathes of ancient DNA have shown that there was a huge migration from this region into western Europe, beginning around 5300 years ago.
The people moving are known to us as the Yamna (in Ukrainian) or Yamnaya (in Russian). In February, I reported on a pair of genetic studies that tried to reconstruct the origins of the Yamna. They concluded that people from the Caucasus moved north-west into what is now Ukraine and mixed with local hunter-gatherers, forming a culture called the Serednii Stih. One subgroup of the Serednii Stih ultimately became the Yamna.
This story had an interesting consequence: it offered a possible explanation for Indo-European languages reaching Anatolia. The same Caucasus groups that contributed to the Yamna may have also contributed to Anatolian peoples.
In short, we have an Indo-European homeland on the steppe north of the Black Sea, but with close connections to the Caucasus and Anatolia to the south and south-west.
This gets me to my point. (Finally, I hear you cry.) In our quixotic endeavour to identify the homeland of the Indo-European languages, how small does the purported homeland need to be before we decide we’ve nailed it? At this point, the argument is between three neighbouring regions: the steppe, the Caucasus and Anatolia. Why does it have to be only one of them? Why couldn’t the homeland be larger?
Remember, we are talking about a time before nation states and hard borders. As the genetics shows, people moved, sometimes over huge distances.
Likewise, Proto-Indo-European didn’t exist in a vacuum. Presumably, there were thousands of other languages being spoken at the time, all jostling up against each other and borrowing words and blurring together when populations mixed.
Mallory discusses whether Proto-Indo-European was a “mixed language”, produced when populations that spoke different languages came together. Many languages contain words borrowed from multiple others: English is famously a confusing mixture of French, Latin, Greek and various Germanic languages. However, Mallory says that a real mixed language is one “where so much of its grammar, not vocabulary, has been spliced together from two languages”, and that these are “extremely rare”.
Still, I find myself questioning how discrete and defined Proto-Indo-European was. To some extent, it depends on how specific our definition is. If we define Proto-Indo-European in some strict way, maybe it was only spoken in a 1000-square-kilometre area of the steppe. But the people in the neighbouring regions may well have spoken a similar dialect.
All of which leads me to suspect that Proto-Indo-European was an emergent phenomenon, one that coalesced over a fairly large area rather than springing fully formed from a single small region. I may well be wrong – but at least I’m not daft enough to claim it originated at the South Pole.
Finally, let’s put all this into perspective. Scholars have spent over 200 years trying to nail down the Indo-European homeland. But there are plenty of other language families. Indo-European has the most speakers, but that’s a legacy of colonialism.
If you look at the number of languages within a family, the biggest appears to be the Atlantic-Congo family, which is spoken across much of Africa. It contains more than three times as many languages as Indo-European.
Yet we know very little about it. If you search for “Indo-European languages” on Google Scholar, you get over 350,000 results. “Atlantic-Congo languages” returns about 1350. It’s just a suggestion, but I think those languages might be worth deeper study.
Embark on a captivating journey through Turkey, a land rich with historical treasures that illuminate the story of human origins.
Human origins: Neolithic and Bronze Age Turkey
