Indo-European Linguistics with Lin-Manuel Rwanda

Proto-Indo-European, Indo-Anatolian, matching genetics with linguistics and more

Linguistics is not my strength, but the field has always been important for understanding prehistory, in particular for how cultures spread and developed. It also brings an entirely separate set of evidence to archaeological interpretations, using methodologies unique to itself. In this way it can help prove, disprove or reinforce hypotheses from material analysis or genetics. Of the major languages in prehistory none has been more influential than Proto-Indo-European (PIE), the ancestor of the modern Indo-European family. The deduction of PIE, and the hunt for its homeland, forms a major theme in world scholarship, launching a thousand ships of research and helping invent archaeology as we know it today.

To walk us through some of these topics I interviewed the brilliant

Lin-Manuel Rwanda

, a good friend of mine. You should all read his substack

Kigali Dispatch

and subscribe to support his work.

---

1) Your work so far is focused on PIE and its evolution in Europe. Can you explain to our readers how you see this process occurring, a narrative for the novices as it were? 

The basics of the "Indo-European Conquest" are something we've broadly known about since the middle of the 19th century, but which were very deliberately forgotten in the latter half of the 20th century as part of an ideological current in academia of which I'm sure even complete novices are now aware. Very briefly: around 5000 years ago, groups of highly aggressive and acquisitive nomads from the steppe regions of eastern Europe and western Asia began a process of very rapid takeover of their neighbours, contributing to the establishment of ethno-linguistic communities from China to Ireland. We call these people "Indo-Europeans" and their language "Proto-Indo-Europeans" because, frankly, we can't agree on a better name. Their descendants include almost all modern European populations and languages and a considerable chunk of the populations and languages of west, central, and southern Asia.

Where I differ somewhat from the mainstream view (that of a series of "wave" migrations from a single culture, usually identified as the Yamnaya or "Pit Grave" archaeological culture of modern day Ukraine) is in the specifics: available philological evidence – bolstered increasingly by advances in population genetics of which your readers will be well aware – suggests that the "Indo-European conquest" occurred in discontinuous stages and was conducted by at least three different but closely related sibling cultures. The most important of these by far is the Corded Ware culture, which was traditionally seen as an offshoot or cadet branch of the Yamnaya culture, but which genetic evidence now suggests was more like its much more aggressive and successful younger brother. 

Almost all Indo-European families (Germanic, Celtic, Italic, Indo-Iranian, Balto-Slavic) are actually "Corded Ware" families, some of which invaded and occupied territories which were formerly home to their "Yamnaya"-derived cousins. (The most interesting example being the Catacomb Culture, of the Pontic Steppe, from which the Greek or "Hellenic" peoples most likely derived, which was partially replaced by early Iranian speakers towards the end of the Bronze Age).

I suspect that this process of back-migration and invasion occurred many times in the history of "Indo-European" as a family, with currently extant "Indo-European" language families having established themselves by conquering and replacing related cultures. At the very least, I think we should suppose the existence of a "ghost" branch in the Balkans which was eventually replaced by the likes of Greek, Albanian, Illyrian, and Daco-Thracian. It may also be the case that other "lost" IE groups were active in the Near Eastern cultural sphere before they were superseded by Indo-Iranians in the 2nd and 1st millennia BC. 

2) Indo-Anatolian is clearly a preoccupation of yours, the first article on your substack is a fairly long breakdown of a recent publication on the topic. Can you summarise the problem here? 

Believe it or not, I actually care very little about "Indo-Anatolian" as a concept. I believe it to be a very alluring red herring against which the evidence is already quite heavily stacked. I only started seriously studying it because one of the theory's most prominent proponents referred me to a paper of his which very neatly summarises all arguments currently available for it after I posted a (fairly lazy and now now deleted) thread about the subject on Twitter. I decided to try and tackle the subject semi-formally because I believe it makes some very fundamental mistakes by trying to mash linguistic and archaeological evidence into a theory of convenience.

Briefly: the Anatolian languages (Hittite, Luwian, and a few other languages spoken in modern day Turkey) are the most divergent Indo-European group, and were extinct by the first centuries AD, replaced by Greek and Latin following the conquests of Alexander and later the Romans. A long time ago, it was proposed that their divergence indicated they were the first Indo-European group to split off from the rest of the family, and that they were so different from the rest they should be considered a completely separate sister-clade of Indo-European proper, originally called "Indo-Hittite" after the most well known member of the Anatolian family. 

This theory was abandoned for many decades before being taken up again with gusto by various researchers at Leiden University (usually called the "Leiden school"), who have gone to heroic lengths to demonstrate it. Many of them believe they can place the divergence of Anatolian from Indo-European at around 1000 years prior to the breakup of the rest of the family into its familiar branches.

My main problem with this idea is that it simply doesn't stand up to common sense. For instance, we can only derive Anatolian words from an Indo-European predecessor by reconstructing a single stage of the protolanguage, with more-or-less identical phonology. Languages don't remain static for a thousand years. Similarly, we can also point to several processes of change at work in other Indo-European languages which were already "active" in Anatolian, and the major differences between the families don't actually require that much time to have developed.

My TLDR on Indo-Anatolian is essentially this: it is real, but it isn't a meaningful category, with Anatolian having broken off early but not too early. If we set the arbitrary date for an Anatolian breakaway at 3000BC or even 3300 or 3500BC, we should see the other branches becoming distinct about 200-300 years afterwards at the latest (although I would personally for reasons I won't go into here place the date slightly later: 2800BC for Anatolian and 2500-ish for the rest). The archaeological evidence of Indo-European settlement in the Balkans and the north and western edge of the Black Sea in the earlier 3rd millennium probably represents a "ghost branch" of the type I posited above, and we shouldn't try to link these to "Proto-Anatolians".

Also, as a personal aesthetic preference, I think "Indo-Anatolian" is both unhelpful (in that it confuses the issue of the relatedness and relations between these languages rather than clarifying it) and profoundly ugly. If we are going to propose new names for things they should at the very least be better than the ones we already have.

3) You've taken on a few fairly challenging words in your writings, trying to understand the origins and construction of terms like 'horse', 'Arthur' and 'Aryan'. Can you speak to any of those and how you go about tackling their etymologies? 

In many cases a lot of the work has already been done, I'm just bridging the gaps between different fields and sub-fields. Prior to the arrival of the internet, the information needed to make these connections was often scattered in specialist grammatical tomes on specific languages (like, say Sanskrit, or Avestan, or Hittite, or more exotic dead languages even than these), the contents of which nobody else has had the inclination or time to synthesise. In the particular case of "Horse", the only thing standing in the way of a coherent theory was the fact that Luwian has been grossly understudied until very recently and the shape of the word for "horse" was similar enough to other languages for people to have made the link there before giving up on the trail. "Aryan" obviously isn't a word many want to spend too much time on for fear that they will be seen as politically suspect. And as for Arthur, the predominant theory on the origin of the name has lead to a chilling effect on the study of alternatives. I only realised where the mistake might have been made through casual conversation with our mutual friend

Aurochs

.

4) We've seen a complete revolution in how the Indo Europeans are understood with the advent of genomics beginning around 2015. How has the linguistic side of IE research responded to this? 

To put it very briefly, it either hasn't (for political reasons), or it has attempted to only in the most trivial terms, either by loosely mapping language movements to population genetics and archaeological evidence but without seriously considering the implications of these on long-standing theories of the family's development. 

For instance, we know that the Afansievo culture of Northeastern Eurasia is closely related to the Yamnaya culture. It is now also assumed as a result to be synonymous with the enigmatic Tocharian language family. Why? Because, erm, Tocharian is Indo-European, the Yamnaya culture is assumed to be the Proto-Indo-Europeans, and Tocharian was spoken fairly close by several thousand years later. In actual fact, Tocharian's unique features and relationships with its neighbours make this connection very unlikely in my opinion (I feel it's much more likely to be part of a later eastward Corded Ware expansion and may in fact have been a component of the famous Sintashta culture, but this is an idea for another time).

I feel that as more evidence becomes available or matures Indo-Europeanists will be forced to radically rethink some of the models that they are currently clinging onto for dear life, and that this will go far beyond the abandonment of the "cultural diffusion" model and into a complete redevelopment of the way we understand Eurasian prehistory. 

5) How does PIE and its study compare to other language families, such as the Bantu languages or Afro-Asiatic? 

In very simple terms, it is much more advanced. Despite many decades of bold attempts, Indo-European represents one of the few "major" language macrofamilies for which we are even able to attempt a coherent internal phylogeny. Although Afro-Asiatic languages are presumed to be related to each other to some degree, there is little agreement on the exact degree of relatedness, and some marginal members of the family are still in dispute. Bantu is particularly underdeveloped thanks to the complex and impenetrable nature of migrations and back migrations common to Sub-Saharan Africa in historical times (in many areas the groups currently present in a given area were likely not present there even a few hundred years ago).

But even in the case of Indo-European, we should always remind ourselves that we know far less than we think we do.