Horses, apples and proto Indo European

After being (quite rightly, it seems) told off by one of my commenters after claiming proto Indo European was 9,000 years old and from Anatolia, I decided to spend a day doing some digging into the subject. Okay Maju; now I get why the Anatolian hypothesis and 9,000 year old date is lame.

Reason 1

First of all there’s the dating. The technologies that date PIE are the wheel, the axle and metal working terms that include gold and white metal (tin or silver). And this rather solidly sets the oldest possible date for PIE anywhere to be 5,500 years old. Anatolia 9,000 years ago is just right out, and I’m embarrassed now that I thought it was correct.

Reason 2

The appearance of the horse and apple, both domesticates from Kazakhstan that hadn’t spread very far by 5,500 years ago, and both are words in PIE. In fact, both domesticates seem to have expanded into Europe and Mesopotamia together, and share an arrival time in Persia with Indo European languages, at about 2,000 BC. The plum also originates from near the Caspian sea, and seems to follow a similar route.

Reason 3

A look at some of the other PIE words showed they had agriculture and a range of domesticated Anatolian/Iranian animals, which eliminates the Botai culture that seems to have domesticated the horse about 5,600 years or more. They had to be in an area about 5,000 to 5,500 years ago that had access to horse, apples and the wheel, and that area of overlap was pretty small, and it didn’t include anywhere West of the Black sea or further East than central Kazakhstan.

Reason 4

The expansion of the IE language group very closely matches the spread of the horse and the apple. IE arrives in Persia about 4,000 years ago, and so do the horse and apple. The arrival of the horse in the Takla Makan also ties with the arrival of the Indo European Tocharians, and it can be seen spreading into Europe, reaching the Mycenaeans about 3,700 years ago, also bringing the horse.

Reason 5

I took a good look at the terms describing the PIE homeland. There  are several words meaning sea, lake and shore, and several for mountain or hill. There are quite a few terms describing trees of various species; yew, beech, willow, birch, fir, ash, oak/hornbeam; all of these describe a fairly cool environment. There’s also a word for snow, and one for ice. Wherever they lived had big lake/seas and boats, as well as mountians. It also knew bears, wolves and and lions (lions used to be seen all through Eurasia and Africa).  There have been attempts to put words like monkey and elephant into PIE, but these seem to be Semitic loan words. Leopard however, may deserve a place, as these are found in mountainous areas in the Caspain/Black Sea area. The flora suggests somewhere cooler than Anatolia.

Reason 6

PIE shares some terms with both proto Semitic and proto Finno Ugric, suggesting it was geographically close to both at one time. The proto Semitic terms it shares are primarily agricultural, like bar (grain), tauro (bull) and waynu (wine), which suggests that the transmission of the words into PIE was pretty early. As a side issue, the placing of goats and sheep in proto Semitic makes an African origin for PS a bit unlikely; as does the transmission of PS words into PIE (which has never been near Africa).

Reason 7

The dates IE languages appear. They appear with the Tocharians 3,800 years ago, in Mycenae 3,700 years ago and in Persia about 4,000 years ago. This suggests a central distribution point somewhere just North of the area between the Black and Caspian seas, assuming the expansion moved at a roughly equal speed in all directions. However, it does narrow down the search area, and it seem to be unlikely to be  anywhere further West than the Black sea, or South of Iran. This likely area is also close to a proposed area for proto Finno- Ugric, the Volga area.


Dates for the appearance of Indo European languages at Mycenae, Takla Makan and Persia. The shaded area is my most likely area of origin solely from the dates and distances. Proto Indo Iranian is thought to date to 4,500 BP in the Northern part of Iran.

A slightly more southerly part of this area, the trans-Caucasus, would have been one of the first areas to have both the wheel and domesticated horses. They also had a mountainous terrain, and access to great quality arsenical copper.


Wheel/Horse area overlap at 5k ago shaded in blue.

This area is also mountainous; and home to willow, birch,yew and hornbeam trees. It even has a leopard native to it (suggested but not proven as a PIE word). The best match I can find for the flora is on the Black Sea coast of the trans-Caucasus area around Krasnodar, so pretty much the area that was picked for the Kurgan hypothesis, just slightly more into the mountainous areas to the South. I’m not  pro the steppes areas in the more Northern possible zone as a homeland, as these wouldn’t account for the plethora of sea/boating related terms, or the trees, or the mountains. These people had plenty of words for mountain and boating, and the steppes, by their nature, are flat, fairly treeless and not easy to sail on.

I’m not sure that the expansion was so much due to direct military conquests as the wheel and horse combo giving them the edge in many areas; agriculture, trade, war… you name it, the cart/chariot has a lot of uses. Wherever Indo Europeans arrive you see horses arrive at the same time.

So what I’m now looking for is a culture dating from 5,500 to 5,000 years ago in the North trans-Caucasian area. There are a few possibilities, but the Maykop culture fits the time and place and geographical/flora and fauna perfectly. I’m investigating them today. So far I haven’t had a good look at the genetic evidence dated to the era, but that’s next on the list of things to check.

I’ve learned a few other things researching this, mainly to do with proto Semitic. These are that:

  • Proto Semitic wasn’t African in origin (I never thought it was, but it’s a nice confirmation), and seems to have radiated out from Anatolia/Iran with the Neolithic expansion, with PIE neighbours.
  • That Elamite (extinct Semitic) is related to Dravidian.
  • Languages can expand almost explosively, and can die out just as quickly.
  • There’s probably a good reason for the Celtic langauges having an Afro Asiatic language structure, which has nothing to do with North Africa. It would seem quite possible that the first farmers who expanded into India and Europe all spoke an AfroAsiataic language, which was then swallowed by IE  (a mirror of what happened in India with Dravidian).

Not a bad day’s research..

14 responses to “Horses, apples and proto Indo European

  1. You saw the light, haleluyah! 😉

    Actually the reasons for my own conviction (apart of the Renfrew hypothesis being generally less echoed through the Academia, i.e. books, papers, etc.) are mostly of archaeological nature: you can in fact trace the Indo-European expansion pretty well in the archaeological record frollowing more or less the sequence outlined by Gimbutas. Gimbutas was maybe a bit too exaggerated on her idealization of pre-IE agricultural societies and demonization of IE nomads, but she nailed the archaeological interpretation very well.

    And you can follow the process almost century by century, if you read on the details of European Chalcolithic and also on the Bronze and Iron age sequences and patterns.

    Some time ago, bored of not finding really good maps on IE expansion in the net (many are very rough approximates, maybe more confusing than explainatory) I created this page that explains how IEs expanded before c.1000 BCE (no mention of the main Iranian and Celtic expansions). Some details may be arguable but it is a good reference, I believe.

    Your review of the linguistic data is also interesting, of course. I read it with interest.

    So what I’m now looking for is a culture dating from 5,000 to 4,500 years ago in the North trans-Caucasian area.

    Well, so far, the mainstream theory actually suggests the Samara valley area a little more to the north/NE (at the Urals in close neighbourhood with proto-Fino-Ugric – check also the Indo-Uralic superfamily/sprachbund theory, nowadays quite consolidated), where archaeology shows a sequence of cultures that appear to lead directly to early Kurgan PIE people, namely the Samara and Khvalynsk cultures, just east of the Volga (western neighbours of the Botai culture and southern ones of proto-Fino-Ugric peoples).

    The Caucasus area was partly (and somewhat confusingly) indoeuropeanized soon after the expansion began (Maikop culture) but it does not seem to be the urheimat at all (and the many pre-IE linguistic pervivences rather add against it too).

    But most important may be the complex Seredny-Stog II horizon (sometimes called Seredny-Stog culture, quite inappropiately – too mixed and plural), in the Don-Dniepr area, where IEs appear to have become more hierarchical, rich and prone to roaming, so to say. The Russo-Ukranian steppe was in any case the starting point of their expansion, that looked westward first of all, albeit in a varied and complex manner. Then (and affecting only the European branches, largely excluding the Balcans, where distinct dynamics took place), Poland-East Germany would become the main node and later, in the Bronze Age, it was the Rhin-Danub area north of the Alps. Expansion to the East (from the steppary core) belongs already to the Bronze Age and it’s contemporary of Celtic expansion in Western Europe.

    Glad that you found the evidence that persuaded you, because, as you say, Renfrew’s hypothesis is “lame” (actually contradicts historical linguistic data is largely just a “rosy legend” that feeds in our romanticing identitarian psycho-emotional needs).

  2. I think all three of us are in agreement here although you’ve both gone more deeply into it than I have.

    I’ve had a big problem with those who claim Indo-European spread with the Neolithic from Anatolia for many of the reasons you list. As for the diversity of the languges being older than 5500 years, there’s what I see as a simple explanation. We would expect the first step in diversification to be development of regional dialects within its limited distribution. This could well have begun as long as 9000 years ago. It’s reasonable to suppose that any group of migrants would have spoken single dialects of any ‘original’ language.

    “A slightly more southerly part of this area, the trans-Caucasus, would have been one of the first areas to have both the wheel and domesticated horses”. I suspect that the two-wheeled chariot was a product of a combined technology, especially seeing that the Hittite region of Anatolia openly included other language groups.

  3. While I’m at it I’ll put in a link to my own thoughts on the subject. I think you’ll find we’re largely in agreement:

  4. what do you think about the theory of Talageri and Elst that PIE came from Kashmir-Bactria ,also a place that coresponde whit recontructed PIE.They also point out south-east asian(like munda) borowings to PIE,thing unlikely if PIE developed in Caspian region.They sugest even an south-east asian origin of PIE.
    Another argument against PIE in the caspian is the fact that caspian region is sparsely populated and so incapable to rise a important army .Is amaizing that peoples from a almost uninhabited region concore half of eurasia and impose their language.
    The argument agaisnt maykop culture as IE is that the region is ‘populated ‘by caucasian languages though they may came later at that extent(remember hurrian-caucasian invasion).

    • I’m favouring a ‘North coast of the Capsian’ location for PIE, which has some support since the wheel seems to have been invented in that area.

  5. what do you think about the theory of Talageri and Elst that PIE came from Kashmir-Bactria ,also a place that coresponde whit recontructed PIE.They also point out south-east asian(like munda) borowings to PIE,thing unlikely if PIE developed in Caspian region.They sugest even an south-east asian origin of PIE.

    Considering that R1a(xR1a1) might well have originated in India, I would be happy if archaeological data would direct pre-proto-IE (pre-Kurgan cultural complex) to that region. But there is nothing so far – though admittedly archaeological research out of Europe is generally in bad shape.

    It is funny that you mention SE Asia because I have been on occasion surprised by Austronesian words (notably numbers 1-4) sounding like IE. I have absolutely no (other) reason to think in such long-distance connection though and would be more confortable if this coincidence (?) would not exist.

    • for me ,is unlikely that austro-asiatic language family has borowing to or from IE.Especialy that numerals are from basic vocabulary and basic vocabulary is resistent to borowing.Another comon word is aku ,in IE ego(i,self) again from basic vocabulary.This send more likely to a very old nostratic origin(pre-proto-IE).But how austo-asiatic have an influence on IE ? Could be 2 posibilites-via Siberia by the Y-chromosome N (this will explain some conections betwin fino-ugric and IE),or via India if R1a originated there.
      Other similarites betwin west Eurasia and austro-asiatics are some textile patters and single piece boaths(monoxyla).
      Or ,acording whit Talageri and Elst ,IE spread from India.
      But the caspian theory didnt answer to this qwestion:how is posible that low-density caspian area to spread its language over 2 heavy-populated places like Europe and India? Thats an enormous achivement for a small group of people.

  6. added:i didnt know that Y-chromosome N (which is so present in russians) have the origin south-east asia 10000-15000 years ago.Indeed pre-protoIE

  7. what about this map?
    the red map show hoe R1a spread from Gujarat-India 16000years ago

  8. My reading of current scholarly research also concurs that the PIE homeland was somewhere between the Black and Caspian Seas, and that the descendents of this general region became mobile around 4300-4000BC, moving south, southwest, and perhaps west over the next millennia. However, there are a couple of gaps in my understanding of the dynamics of the colonisation of the western Black Sea lands.

    1 Most of the studies seem to focus on land migration, when it would appear that marine technology of the day would have enabled substantial travel across the Black Sea, as well as around it.
    2 It is unclear to me where the farmers of the Danube delta went following the expansion of the PIE culture. Did they become refugees, moving into the nearby hilly regions to the west and northwest, and/or take to the Sea and return to their southern homelands, or the Aegean/Mediterranean archipelago?

    I would be most interested on your thoughts about these apparent gaps in the story.

  9. First off. Good to have you back in action Mathilda.

    “it would appear that marine technology of the day would have enabled substantial travel across the Black Sea, as well as around it”.


    “It is unclear to me where the farmers of the Danube delta went following the expansion of the PIE culture. Did they become refugees, moving into the nearby hilly regions to the west and northwest”

    I think so. The distribution of Y-hap I in Europe suggests a population pushed into such regions by some later expansion, perhaps Y-hap R1a. This would place Y-hap R1b in Europe before Y-hap I. But Y-hap R1b has, in turn, become largely confined to the far western margin through subsequent Y-hap immigration into Europe from the east.

  10. Good research. This is probably the most convincing collection of evidence for a Kurgan as opposed to an Anatolian origin for PIE that I have seen yet.

    I’d be curious about what you think of Bernard Sergent’s Afro-Dravidian hypothesis.

    • Bernard Sergent’s Afro-Dravidian hypothesis.

      I’m of the opinion that there’s a sporting chance Afro Asiatic and Dravidian both had a common ancestor in the Iran/Syria are about 11k ago, as theres an expansion from this area into Pakistan/India as well as North Africa. Otherwise , not really my area.

  11. Your points about age depth are on point. The geographical origins of the apple may be closer to Eastern Turkey.

    An alternate place where you could get all the technologies together at the right time is mature Harappan. They had signficiant trade with Central Asia (providing a source for the technologies there) and had maritime trade — there are many common maritime words in PIE. There isn’t evidence for very sophisticated martime activity by the Pontic-Caspian plains people. And, they also had a reason to burst onto the world scene when they did. The river system that supported most of their cities (the Savasvati) dried up right around the time of IE expansion forcing them to seek a new homeland. The drying of the river can also explain the aversion to burial characteristic of IE expansion, a transition that is motivated if there is mass death from a drying river making burial impractical and also changing the disease equation with aridity.

    If you don’t have a Harappan PIE homeland, you have to explain the presence of farmer associated but not herder associated haplotypes in the IE associated population genetic package (e.g. the low J1/J2 ratio, which is the reverse in the North Caucuses) and you have to explain the very complete adoption of North India specific Vedic material that might not have been relevant to invaders. Uralic which expanded from a similar region around then has lots of Y-DNA N haplotypes associated with pastoralism.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s