Indo-European language spread (part 1)

Archaeogenetics is rapidly rewriting many of the details of what we thought about history and para-history and pre-history, although if you take a step back from the details, what is also quite notable is that it is confirming that much of what the archaeologists, anthropologists and linguists said a century ago was right all along. Even though subsequent work, which in hindsight is now clearly exhibiting socio-political bias of a type that can only fairly be termed "proto-Woke" has been working hard to erase those theories for decades. And it is still important to approach this from a multidisciplinary angle. If genetic data is saying something very different than archaeology, linguistics and physical anthropology, then let's be careful not to assume that archaeogenetics trumps other lines of evidence. The best and most likely true narratives of the past are those in which data from many disciplines all converge on a single null hypothesis.

That said, where are we right now with regards to the formation of Europe, and specifically the prehistoric spread of the Indo-European languages? Because let's face it; although there has been some jostling here and there (like the spread of Anglo-Saxon over Celtic Britain, for example) the reality is that from a cultural and genetic perspective, the Anglo-Saxons and the Celtic Britons were already very similar, and once the Indo-Europeanization of Europe happened in the Stone Age and early Bronze Age, that really truly was the formation of a modern Europe that culturally, linguistically and genetically resembles Europe of today—minus the ridiculous influx of middle eastern and African and Asian DNA and culture that is the hallmark of the very last few years of Europe's history. That isn't to say that there haven't been significant changes here and there; the intrusion of the Turks into our sphere (and I'm American, but that means that my heritage is European, and I consider Europe to me part of "our" civilization) being a notable example, but by and large, the Indo-Europeanization of Europe was the creation of Europe as we know it. And because it happened in the prehistoric era, then Europe has always been known to us in the form that it is now. More or less.

By and large, the earliest peopling of Europe by anatomically modern (i.e., not Neanderthal) people that has an impact on today's population was the spread of people from Ice Age refugia back into the now glacially free territory of the continent. I personally believe that any thoughts on where the people who made up the refugia in the first place came from to be particularly handwavy and not very useful, so I'll skip ahead. 

As one might expect, being confined by climate to small and disconnected refugia, the genetic signatures tend to run in clusters that are significantly different to each other. One of the post-glacial population clusters that persisted through the Mesolithic and which was still important (although probably significantly admixed) in the Neolithic is the Western Hunter Gatherer, or WHG cluster. The origins of the WHG can be modeled by blending the old Ice Age El Mirón cluster from the Franco-Cantabrian refuge and Ancient North Eurasian from the Mal'ta cluster (a much more distant Siberian/Lake Baikal area refuge). Some classic examples of skeletons that belong to this genetic cluster include Cheddar Man, La Brana man, and Loschbour man. There's some controversy which remains unresolved about the phenotype of the WHG, although it's certainly possible that because nobody alive today has a preponderance of this genetic cluster, that they don't resemble anyone alive today. The "classic" image of the WHG is that he has brown skin and hair, blue eyes, high cheekbones and a wide face, and a rather stocky, robust build. However, some studies suggest that lighter skin swept through Europe as a breeding bias event at some point, maybe as early as the middle to early Mesolithic. (As an aside, the highest concentration today of WHG ancestry corresponds to peoples of the Baltic, which are among the whitest on Earth.) While the WHG cluster seems to have been prevalent through much of western, southern, and central Europe, as well as making up a significant element in Fennoscandia as well, it does seem to have some distinguishable structure. The I2 paternal haplogroup, for instance, is most heavily associated with the Balkan Dinaric population, and that haplogroup falls off rapidly as one enters central Europe heading towards either Northern or Western Europe, while it does seem to have spread northeastward, probably with the heavily WHG-centered Narva and Kunda cultures. Northern Europe, on the other hand, seems to be strongly associated with the I1 haplogroup, which is still a major Germanic genetic marker today (although they obviously didn't speak Germanic in the Mesolithic.) Although it's important to keep in mind that the prevalence of a haplogroup can spread rapidly without altering the autosomal sequence very much, i.e., one needs to be careful about assuming that a haplogroup spread means a change in population; it often does not. It can have much more to do with the success of a single individual and his family, sometimes, within a continuous population.

Today, almost everyone in Europe shares at least some WHG genetics, although no population is majority WHG anymore. No doubt the spread of the WHG peoples from the various Ice Age refugia had languages associated with them, but we have no idea what languages they may have been, and if any non-Indo-European language spoken today or in the historical period may have belonged to them. Given the El Mirón refugia played a role in the spread of the WHG genetic cluster, and that's smack dab in the middle of Basque country, there is at least a handwavy connection based on geography that might suggest that Basque is a lingering WHG language. But take that with not just a grain of salt, but a whole freakin' bag. After all, the Balkans and Italy were probably connected during the LGM, and the Villabruna cluster is another major contributor to pre-WHG Europe. By the same logic that would call Basque a possible WHG language, we could call Etruscan one as well. Both are speculative handwavy efforts either way, though.

Speaking of which, there's some potentially interesting structure going on with three of the five major refugia genetic clusters and the associated archaeological cultures. But like I said, while the WHG population didn't spring from the Earth out of nowhere, I think the details of Paleolithic migration and admixing that led to the formation of the group is difficult to recover with a lot of confidence, at least until more and better data is available. The Villabruna cluster seems to be associated with a rare and early R1b (V88) haplogroup in Europe, and may have also used a "Green Sahara" corridor to spread the haplogroup to the Chad Lake area of Africa, where it sticks out like a somewhat confusing and odd discontinuity. Nobody thinks of Europeans colonizing Africa in the Stone Age, but apparently that did, in fact, happen.

Anyhoo, the next Mesolithic population cluster we need to discuss is the Eastern Hunter Gatherers (EHG). This group was distinct from the WHG group, and formed a pretty hard and fast geographic boundary from the western Black Sea and lower reaches of the Danube River, up along the Dniester valley, to the western Baltic, with the only discontinuity being that a blob of WHG genetics seems to have penetrated the line and been more prevalent in the Eastern Baltic, at least for a time (the aforementioned Kunda and Narva cultures.) This didn't last forever, as the later Pit-Comb Culture, which had a much higher percentage of EHG genetics. The EHGs were predominantly Ancient North Eurasian (from the Mal'ta LGM cluster) with some degree of admixture from a population that while not exactly WHG, was probably descended from the same ancestors as WHG; part of the old Paleolithic Gravettian culture, most likely. The genetics for blue eyes are not as prominent among the EHG as among the WHG, so they were more likely to have had brown eyes. There isn't any controversy about skin color as with the WHG, and they are presumed to have been light skinned. Whether or not blondism had spread through this population or not is still an open question as well, as some of them seem to have had alleles that suggest it, and the case is growing over time that it was present among them. The haplogroups associated with the WHG group seem to be heavily biased towards I1 and I2, whereas the EHG favors haplogroups R1a and R1b. The spread of this haplogroup to becoming the most common in Europe (and elsewhere) today are closely associated with the spread of Indo-European languages, although of course, it's nonsense to talk about Indo-European existing in the Mesolithic. There is, however, a better than handwavy reason to believe that the ancient ancestor of the Proto-Indo-European language was originally native to an EHG group. By the time we get to the Neolithic and later, there are no groups that truly can be called an "EHG" population, although certainly some groups had high percentages of EHG genetics still.

The Scandinavian Hunter Gatherers (SHG) are the only Mesolithic group that appears to be a hybrid. Now, don't get me wrong; along the contact zone between WHG and EHG there are a number of cultures that show signs of admixture between the two, but the SHG group is one that hybridized very early and then remained in place, subject to its own genetic drift separate from trends going on specifically in the greater WHG and EHG populations. Also, pretty unadmixed SHG cultures remained in Scandinavia into the late Neolithic or even Chalcolithic or Early Bronze Age, like the Pitted Ware culture that coexisted without significant admixture with the Funnelbeaker culture and later with the Battle Axe culture (and even far enough south to border the Single Grave culture), before finally being absorbed by them as recently as 2300 BC. In any case, the origin of the SHGs is presumed to be WHG migration up the Danish peninsula and across into southern Sweden/Norway meeting EHGs who were migrating southward along the Norwegian coastline. These two populations merged and then displayed a remarkably long continuity. Phenotypically, the SHGs are believed to have already developed high percentages of blue eyes, light skin and blonde hair. Today, modern Scandinavians have a notable degree of SHG still, from the Pitted Ware Culture, presumably, which was their last hurrah before merging with other groups to become something much more similar to modern Scandinavians. The Lapplander people to their north do not have any of this ancestry, and seem to be a much more recent intrusion into the area. This is a reversal of what we thought even just a relatively few years ago, where the Lapplanders and other Finno-Ugric groups of Scandinavia were assumed to be the autochthonous people of the peninsula. Incorrectly, as it turns out, as the intrusion of their languages is more recent even than the Indo-European spread, and is associated with genetic haplogroups originating far to the east in Siberia. 

The Caucasian Hunter Gatherers is the next group we need to discuss (CHG) associated with the Satsurblia cluster from the Upper Paleolithic and the Kotias Cave from the Mesolithic. Apparently while the Caucasus Mountains were themselves likely glaciated, the area in between and/or around them was one of the population refugia, associated with the J haplogroup (itself a descendent of IJ.) Because of this, the CHG are assumed to have been related to the ancestors of the WHG as well from some point in the Paleolithic, although they had a long degree of separation between them. Either less work has been done on this group, or I'm just not aware of it and unable to find it, but there doesn't seem to be good info on the phenotype alelles of the CHG population, nor really good fits to them and modern Caucasian populations, although presumably some of the native Caucasian groups that are unrelated to the Indo-Europeans or other groups had their genesis somewhere with this group. CHGs make up a significant portion of the Western Steppe Herders population group that seems to be mostly although not completely made up of EHG male lines and CHG female lines. (But more on that in part 2). There's also a linguistic theory by Bomhardt that mimics this genetic evidence, suggesting that early Indo-European developed as a Eurasian EHG language with a Caucasian (CHG) linguistic substrate, which pulled the character of primordial pre-Indo-European into certain correspondences and resemblances with a primordial pre-Northwest Caucasian (Abkhaz and Circassian languages, among others.) Of course, plenty was going on throughout history in the Caucasus, and linking the CHG to various later attested cultural and/or linguistic groups, to say nothing to currently attested cultural and/or linguistic groups, is a tricky proposition. Figuring out the tangle of the culturo-linguistic situation in the Caucasus is extraordinarily tricky. And clearly a number of peoples have come and gone, bringing with them a number of languages. Where does Kura-Araxes fit into the picture, or Maikop? How does Hurrian or Hattic relate to any languages that we know from other sources? Etc. See at least one proposal at the end of this post, however.

The final group is usually called Early European Farmers (EEF), and are strongly associated with the spread of Neolithic farming technology. But they weren't just teachers of new ways of making a living; the spread of the Neolithic economy happened with the spread of new people with new genetics, and there was relatively little mixing (at the Mesolithic to Neolithic transition, at least) of these newcomers with the hunter gatherers that were already there. For that matter, I'm talking about the eve of the expansion of Neolithic farmers, so this group maybe should be called Anatolian Hunter Gatherers still at this stage (AHG) although technically that term is usually used for an Epipaleolithic group, although to be fair, they are the same people. The map below shows them as stretching into the Levant, but that's an older, and out of date map—the AHG/EEF were not the same people as the Neolithic peoples of the Levant like the Natufians. (It's got other problems too, but they're minor compared to what it does offer. Even if a bit out of date, it's the only map I know of that actually shows these stocks more or less correct. Albeit without SHG.) Their ancestry reflects deep, Paleolithic splits from the CHG and WHG populations, as well as paleolithic gene flow of Iron Gates and Villabruna Europeans into Anatolia. I know, of course, that Anatolia is considered part of Asia—in fact, the very name Asia comes from a Greek name for Anatolia probably based on the Assuwa Confederation—and while Anatolia is in some ways a cross-roads of the Near East and Europe, for the most part, Anatolia has been culturally and genetically and linguistically more closely associated with Europe than with the Near East, at least up until the Moslem conquests, the destruction of the Byzantine Greek Christians, and ultimately the arrival of the Ottoman Turks. The EEF peoples seem to be shorter and more gracile than the European hunter gatherers. Their phenotype otherwise seems to be a bit unclear. There are hints of blue eyes, blond hair and light skin among them from some studies, and yet all of the places where their genetics is strongest are "Mediterranean" in appearance; darker skin, dark hair, and brown eyes; like most Spaniards, Greeks or Italians. The Sardinians in particular are the population today that has the highest percentage of EEF ancestry, to the point that they can almost be called a relict EEF population still, and if you've ever been to Sardinia, you'll notice that tall, blonde, blue-eyed Nordic looking Europeans aren't really consistent with the natives. For now, let's keep the EEF in Anatolia, but we'll note that as they spread throughout Europe, they will dramatically change the cultural and genetic footprint of the continent—and almost certainly its linguistic footprint as well, although nobody knows what languages the EEF people brought with them.

Let's have a look at that, shall we, very briefly. A lot of proposals over the years presume that the Anatolian hunter gatherers come from the Levant and therefore have Middle Eastern languages, like an early Afro-Asiatic language. They, then, presume that the European language of the EEF cultures was probably Afro-Asiatic. This isn't actually correct, although it's not impossible, it's unlikely that the Anatolians spoke an Afro-Asiatic language, or were very closely related to the Levantine early agriculture. What genetic studies we do have suggest that the Anatolian Neolithic was the adoption of farming techniques by the locals rather than demic diffusion of Levantine/Natufian farmers into Anatolia. Let's first look at what the genetic clusters are, which most likely had unique languages at one point in time, and then look at what historically attested languages we know of that are non-Indo-European, and then see if we can, by working from both ends, narrow down possibilities.

There are six genetic clusters from the LGM that presumably had wholly unique languages. These clusters/stocks may have split as well, and given the time depth, it would be hard to recover any hint of that in anything surviving today. These clusters are:

1. The Franco-Cantabrian El Mirón cluster associated with the Magdalenian culture, smack dab in the middle of what is Basque country today. (19-14k years ago.) This lineage is also associated with an even older paleolithic lineage called Goyet. It seems to have shrunk in area at some point, and then re-expanded. Magdalenian supercedes the Solutrean in time and space, although I'm not sure what—if any—relationship has been posited between the two.
2. The Villabruna cluster (14k-7k years ago) from northern Italy, which spread across the Adriatic into the Balkans as well. Associated with the Epigravettian culture.
3. The Vestonice cluster (34-26k years ago), which is older than any lineages except the Goyet culture (see above) and is associated with what are extremely rare haplogroups today, like CT and with the Gravettian culture. Given its extreme age, I'm not sure what actually happened to it and if there's really any relevance to it from a linguistic standpoint; i.e., could it be the source of a linguistic stock that survived into the eve of the historical era? Probably not.
4. The Mal'ta cluster (24-17k) associated with the Mal'ta Buret' culture and the Ancient North Eurasian genetic profile. In spite of its age, we can actually draw a likely line between this and languages that we know of today. The ANE genetics makes up 14-38% (depending on the specific population) of the genetics of the Injuns, for instance, and significant portions of the Siberians. The EHG cluster, mentioned above, is highly ANE; up to 90+% in some studies, and more like 75% in others. (They also make up about a third of the CHG ancestry, however.) This cluster no doubt had its own language at one time, but given its extreme age, it's the more derived languages that EHG peoples may have spoken, descended over many, many generations from a more pure ANE cluster, that command our attention. Handwavy though the idea may be, it is usually assumed as likely that the most ancient earliest form of something that could be considered an immediate ancestor of proto-Indo-European is a descendant of some kind of EHG language that ultimately came from this refugia.
5. The Satsurblia cluster (13k-10k) is directly associated with the CHG population. Here there is a lot of action for potential linguistic origins; there is a popular (albeit by no means universally accepted) theory that a CHG substrate applied to an EHG superstrate is what caused the Indo-European languages to look the way that they did in their earliest form. Certainly as the groups formed that later emerged as the first speakers of Indo-European, EHG and CHG genetics make up by far the most prevalent genetic profiles (more on that in part 2). It is also believed likely that one or more of the unique language families of the Caucasus is descended from a CHG language as well, but that's just because of the convenience of geography and no other alternatives that are better rather than because of any particularly strong evidence.
6. The Anatolian Farmers (8.3-6k) who are the descendants of the earlier Anatolian Hunter Gatherers and the immediate ancestors of the EEF group is the last cluster. The seem to have split from the WHG a long time ago (~45k) and from the CHG also a long time ago although not nearly as much so (~25k). They also cluster on PCAs in an intermediate position between WHGs and Natufian Levantine farmers—although this isn't meant to imply that they had Natufian ancestry. That's been claimed in the past, but more recent studies seem to suggest that the Anatolian farming revolution happened without demic diffusion from the Middle East, as mentioned above. In any case, the single point of origin of all of the Neolithic farmers would imply that at least initially they probably spoke a closely related cluster of languages, although as they spread through the continent, diverged into multiple archaeological cultures and remained separated from each other for thousands of years, that was probably no longer the case that they all spoke languages with easily discernable similarities anymore by the time the Indo-Europeans arrived on the scene. That said, easily discernable and "recoverable if specialists actually had the data" are two different things, and it's likely that the EEF language family was at least a single family, albeit one that would have branched out into multiple distinct languages.
7. This last one is a bonus, and I'm just throwing it out there not because we have any evidence for it, but because we can't rule it out. There was probably some gene flow between the Maghreb and Europe, i.e., the part of North Africa that's just on the other side of southern Spain. Given that sea levels were lower during the Ice Age, if there wasn't an actual land bridge connecting the two, then it would still have been visible from one coast to another at certain points, and it's unlikely that some north African gene flow didn't make its way into southern Europe. We're already talking about an R1b-V88 flow from the Villabruna, through the Green Sahara and to prehistoric Lake Mega Chad (that's really its name; I know. If I ever decide to change my online pen name, Mega Chad has gotta be an option, although not because of the African connection, which I don't care about at all.) There's no reason to assume that the reverse couldn't have happened. So again, although we have no direct evidence for this, because we can't rule it out, the arrival of yet another completely separate language family from North Africa into the Paleolithic, Mesolithic or Neolithic isn't impossible and can't said not to have happened for sure.

So, now that we've determined a few potential fonts for completely unrelated language families prior to the Indo-Europeanization of Europe, let's talk about what we know about languages that are or were in Europe that are non-Indo-European.

1. Uralic. Uralic is probably the most obvious, being that the national languages of Hungary, Estonian, Finland, and the Lapplanders and some other minor tribes of northern Russia all belong to this family. That said, it does not quality nor will it lead us anywhere. While it was once thought that it was the native, pre-Indo-European language of northern Europe (minus Hungarian, who's arrival in the Middle Ages is a matter of historical record). The Uralic languages can now confidently be traced as a relatively late arrival to the Baltic and Fennoscandia region; no earlier than the Iron Age, in fact. A genetic haplogroup is strongly associated with the spread of Uralic languages, N1C-Tat, and its origins is in western Siberia, or thereabouts. There are some potential links to even more ancient populations, such as the Keltiminar culture of the Karakum desert, the even older Botai culture (of more or less the same region) and more probably, the Seima-Turbino archaeological horizon, which may actually represent the spread of the languages. That said, in northern Europe, the Uralic speaking people obviously don't have a very Siberian phenotype anymore, and autosomally are similar to other Scandinavian, Baltic and Slavic peoples who are their neighbors; the Hungarians are genetically identical to their neighbors as well. We'll have to look elsewhere for pre-Indo-European languages of Europe.
2. Basque. The other extant non-Indo-European native language is Basque. It has been spoken in the Cantabrian region, the Pyrenees, and nearby Gascony since the beginning of recorded history (Gascony is actually named for this people) and is related to the historically attested Aquitanian language. Whether Aquitanian is a direct ancestor or a sister language to paleo-Basque of some kind if unknown, but probably more likely the former. While no evidence has ever amounted to a convincing connection between Basque and any other language, it is likely that it's a lingering remnant of a group of languages that was once more widespread. Whether this is full-blown Vasconic Substrate Theory of Theo Vennemen or something more constrained is unknown. Vennemen's theory doesn't have widespread acceptance, but not because anybody thinks that it's unlikely really that there was a Vasconic substrate to parts of the Atlantic coast of Western Europe—just that there isn't any evidence to prove it in the format that he's presented it. If it is a remnant EEF language, then it's pretty compelling evidence that the EEF's didn't bring anything at all similar to Afro-Asiatic, but then again, nobody knows who originally brought Basque to Europe. Given that it overlaps geographically with the El Mirón refugia, I kind of like the idea that it may be the last evidence of the languages spoken by the El Mirón genetic cluster; stretching all the way back to the Ice Age. If that's true, then the "golden age" of Vasconic territorial expanse would have been way back in the Aurignacian culture of Europe. But that's a major handwavy bit of speculation. 
3. While we're in the same general vicinity, two Paleohispanic languages other than Basque and various Indo-European tongues are known, but completely unclassified, or only classified in a very controversial way: Tartessian and Iberian. Some scholars prefer to see Tartessian as an Indo-European language, or even a specifically Celtiberian language, but the majority suggest that it cannot be so interpreted with confidence, and there are good reasons to reject it. They would suggest that Tartessian is probably a language isolate, although the script is sometimes seen as compatible with a Basque or Iberian-like structure. This means that Tartessian might be Indo-European, a Vasconic language, or something else unique. Unfortunately, none of that is very usable in interpreting the pre-Indo-European landscape of Europe. Iberian, on the other hand, "feels" more likely to be linked with Basque, although hardly anyone would assert that we can say that with confidence. Rather, it's at best a "this could be likely, but we don't know" kind of situation. Personally, I kind of like the idea of both Iberian and Tartessian being part of a Vasconic family. We also know that later in the Neolithic, WHG genetics increased across Europe due to demic diffusion of new cultures like Globular Amphora, which had a very elevated percentage of WHG DNA compared to the cultures that they replaced, and that it seems to have come from the west. Of course, genes don't speak languages exactly—although we do find that language replacement does tend to coincide with an influx of new genetic markers even so. I kind of like the idea of all three of these being Vasconic, and originating with the old WHG population of the peninsula, but I admit that that's just a pet bit of speculation that I would never propose seriously. Keep in mind that even in historical records we're aware of plenty of other peoples in the peninsula while the early Greeks and Romans were making records of it; Hispania was a stronghold of the Carthiginians before the end of the Punic Wars, and presumably plenty of people from North Africa or even the Levant could have had an imprint on the peninsula before it was even Latinized. Also; by the time the Greeks and Romans entered the picture, Iberia had been largely Celticised as well; Celtiberian tribes being widespread across the territory, and they had been there since the early Bronze Age expansion of Bell Beakers across whatever was there before that.
4. We don't know too much about the Italic peninsula linguistically prior to the arrival of the Indo-Europeans with the major exception of the Etruscan language, which is deciphered and can be read. It's possible that this is also a language family of once broader distribution; Rhaetic in the Alps, Lemnian in the Aegean and a few others (like Calumnian and North Picene) are proposed as possible members of this family, although it's doubtful that we'll ever have enough material to say that confidently. That said, linguists are often fond of throwing ideas out there without necessarily having the confidence that they should, so this so-called Tyrsenian language family is often proposed. Some even link it with the Aegean languages (see below) which may or may not be a family in their own right. In my opinion, Etruscan is most likely seen as an ancient EEF language that spread through the initial Cardial Ware expansion, diversified in southern Europe, and was still kicking around to be documented by literate civilizations like the Greeks, Romans and Phoenicians. Although I also think that the Aegean languages (again, see below) are a good candidate for that, and given that there is little reason to see those two groups as related, most likely only one of the two is true. Now, this isn't necessarily true. Etruscans might well be intrusive from further north. Ancient authors offered several hypotheses: Dionysius of Halikarnassus suggests that the Etruscans were truly autochthonous, but Herodotus and a few other Greek historians suggested that they were related to the Pelasgians and had migrated to Etruria from the Aegean. Roman historians Livy and Pliny the Elder—who may have had more first-hand experience with the Etruscans than the Greeks did—believed that their origin was in the North. DNA evidence suggests that they are genetically extremely similar to the Latins and has most in common with Central European Urnfield peoples, including a fair bit of steppe-related ancestry, so it's no help either. Maybe if we knew a bit more, we could assert that Etruscan was a descendent of the northern expansion of the EEF rather than the southern one, and was intrusive into the Italian sphere. That way we could salvage an eastern Mediterranean and Aegean group of languages associated with the Cardian Ware expansion along the northern coastline of the Med, while Etruscan might represent a long-separated EEF language of the other fork that went through the Balkans to Central Europe all the way up to lower Scandinavia. While ultimately originating from the same source, their separation since the very beginning of the Neolithic might well make that common ancestry difficult if not impossible to spot. Of course, all of this is just speculation.
5. OK, I've referred to it a number of times; time for the entrance of the Aegean languages. On the other hand, although we certainly know that they existed, we know, sadly, very little about any of them. The Greeks had a long habit of referring to the Pelasgians as an autochthonous people of Greece who were nevertheless very notably not Greeks. However, they seem to have made no effort to document much if anything about their language, so theories run wild as to what Pelasgian was. There's not even any widespread agreement that it was non-Indo-European, given that our models suggest Anatolian languages spread over the Balkans and other nearby areas before the Greeks arrived, and were still speaking their languages just on the eastern and northern edges of the Aegean into the historical era. The reality is that even though many of these peoples are known to us from history, and maybe even some of their kings and some of the things that they did, we don't know much at all about them culturally or linguistically, so identified who they were related to is difficult. The Assuwa League, including Troy itself, likely spoke some kind of Anatolian Indo-European, and Lycian/Lydian branches of the same family ran down the coast further south, while Luwian and Hittite itself were more in the interior and further east. Regardless of the difficulties with Pelasgian or any other specifically Aegean language, one nearby language that is known for sure to be non-Indo-European, and to have been the native tongue of a successful maritime trading civilization is Minoan, and it's possible descendent Eteocretan, both of which on the island of Crete. Their language is attested, but untranslatable, and other than the fact that it's not related to anything that we know if, we don't know much about it. I think it likely that Minoan along with Etruscan represent our best bets at recovering a late EEF language and culture that we can talk about in terms other than about their pots, cemeteries and the ruins of their houses and villages.
6. If the EEF peoples originally came from Anatolia in the early Neolithic, then the likely place to look for languages that might be related to theirs would, naturally by the non-Indo-European languages of Anatolia. Duh. As it happens, Anatolia is mostly known for its various Indo-European languages from the Anatolian family, to later appearing Phrygian, Greek, Latin and others. The Semitic language of the Assyrian and later Aramean traders were obviously intrusive (and late, at that) in small doses from the south. That said, there are three languages or language families that demand our attention, starting first with Hattic. Hattic has the same place among the Hittites that Pelasgian has among the Greeks; the supposed language of the autochthonous people. A little bit more is known about it than we know of Pelasgian, but certainly not enough that anyone could write a Hattic text or anything like that. The conservative view is that Hattic is simply another isolate, but there are attempts to connect it various other nearby languages, including the Northeast Caucasian languages (mentioned above in the CHG section; sometimes called the Circassian languages) or the Kartvelian language family (the family to which Georgian belongs; Stalin's native language.) If anything comes of those proposed connections, which seems unlikely, then it makes it more likely that Hattic is an example of spreading from the Caucasus, and would make it less likely to be related to an EEF language. However, if it turned out that the EEF and CHG languages actually had a very distant genetic link, that would sure be something else, wouldn't it?
7. Kaskian is another non-Indo-European language of the southern Black Sea shore. Although very little is known of it, it's been suggested that it might be related to Hattic, and that it might be related to Circassian. In fact, it's been suggested that Kaskian and Circassian are reflexes of the same word over time. 
8. A third major Anatolian language known from the Bronze Age is Hurrian and the closely related Urartian language. The Hurrians seem to have been kicking around in the region for some time, and it's likely that the Kura-Araxes culture is a proto-Hurrian speaking culture, and that they may well have been early traders of viticulture, or even the inventors of it. There are suggestions that their language, which is known, by the way, has connections to the Northeast Caucasian languages, i.e., the family which contains Chechyan, for example. There are also genetic links between the two groups/areas, which makes this even more likely

Although these aren't linguistic theories, there are some new genetic theories that might have bearing on linguistics, even if they can't be proven. One noted here, suggests that there was a third major, expansionary family from the general area; not Indo-European, not Semitic but something else entirely. That paper proposes as a hypothesis in need of more work that a family consisting of Pre-Greek (Pelasgian), Minoan, Hattic, Hurro-Urartian and both the Northwest and Northeast Caucasian languages (which some linguists link together with each other, although other more conservative ones do not.) If this language family, which has no label that I'm aware of, has any currency at all, then it's a prime candidate for the language family of the Anatolian Farmers and their migratory brethren, the EEF people. There are hints that the pre-Greek substrate has some similarities to pre-Germanic substrates, which would actually bolster that idea; the idea of a Hatto-Hurrian-Pelasgian-Minoan-etc. language family being the EEF language family that spread across most of Europe with the spread of the clearly related Neolithic invasion of Europe from Anatolia makes all kinds of sense.

This likely leaves Etruscan/Tyrsenian and Vasconic as unrelated to this EEF language family, making it slightly more likely that their roots go back to the Mesolithic WHG peoples and whatever languages they may have spoken.

It's also important to keep in mind that there could well have been more than one substrate in Pre-Greek; there's evidence to suggest gene flow from the Chalcolithic Iranian highlands south of the Caspian Sea, for instance. And while I've said I don't favor an Afro-Asiatic affiliation as being the affiliation for EEF languages, it is clear that at least some likely Afro-Asiatic peoples interacted with the EEF peoples of Anatolia, the Aegean and Crete, Cyprus, etc. 

Anyway, more questions than answers here, but we at least have put most of the cards on the table (all that will fit in a reasonably long blog post, anyway) of what we know of the linguistic/cultural situation of Europe before the Indo-European languages erupt on it. In part 2 I'll talk a bit about the cultures of EEF Europe and the genesis within the Neolithic of the Indo-Europeans, before we talk about how they spread, which will probably end up being a part 3.