What is the Homeland Problem?
In my recent discussions with blog readers and students, it has become clear to me that before considering the Proto-Indo-European (PIE) homeland problem in more detail it might be beneficial to review what a linguistic homeland is in the first place. Particularly at issue is the possibility of a “compromise” theory: given that both the Steppe and the Anatolian theories of the PIE homeland have their proponents that put forward certain arguments and provide some evidence for their claims, could it be that both of these theories are correct and that PIE homeland is located in both southern Russia and Turkey? For example, my former student and co-author of a paper published in Science, Rory van Tuyl writes:
“Haak [et al. 2015] shows that Europeans are genetically admixed between Early Neolithic peoples (presumably from Anatolia) and late Neolithic peoples (from the Pontic steppes). Why can’t we assume that the modern IE languages of Europe have origins that are also admixtures?”
Another reader writes on Facebook (spellings corrected):
“Personally, I’m skeptical of any uni-variant migration story for IE languages. People move in waves and they move in complicated fashions”
A similar compromise theory is explored in Piazza and Cavalli-Sforza (2006), though they end up siding with the Steppe theory:
“…if the expansions began at 9,500 years ago from Anatolia and at 6,000 years ago from the Yamnaya culture region, then a 3,500-year period elapsed during their migration to the Volga-Don region from Anatolia, probably through the Balkans. There a completely new, mostly pastoral culture developed under the stimulus of an environment unfavorable to standard agriculture, but offering new attractive possibilities. Our hypothesis is, therefore, that Indo-European languages derived from a secondary expansion from the Yamnaya culture region after the Neolithic farmers, possibly coming from Anatolia and settled there, developing pastoral nomadism.”
However, a compromise theory of this sort is impossible, by definition and here is why. The linguistic homeland of an ancestral language is defined as the last place where that ancestral language was spoken in the time leading up to its first split. In the case of PIE, most scholars agree that the first split was between the Anatolian branch and the rest of the family (dubbed, for ease of reference, Proto-Nuclear-Indo-European, or PNIE). Thus, the PIE homeland is, by definition, the place where it was spoken in the time leading up to the separation of the Anatolian languages. Where was that location and could it have been more than one place?
Before considering this issue, let me illustrate with a much more recent and hence much better-understood example. As a little thought experiment, imagine a Martian who has landed on Earth and knows nothing about human history, but only about the current state of affairs. For whatever reason, the Martian gets particularly curious about Portuguese, noticing that there are two varieties of it: Brazilian and European. The Ethnologue, which tends to err on the side of splitting rather than lumping, surprisingly treats the two as dialects of the same language. Yet many other sources treat them as different languages. Linguistically speaking, European and Brazilian Portuguese are quite distinctive, especially in the spoken/colloquial language, more so than British and American English (which, by the way, some people consider different languages). Brazilian Portuguese has many words absent in the European variety, but importantly the differences between them go well beyond words. For example, European Portuguese allows null subjects and verb-subject orders, whereas Brazilian Portuguese does not. The progressive aspect is expressed in European Portuguese by estar a + infinitive (e.g. Ela está a dançar ‘She is dancing’), whereas in Brazilian Portuguese it is expressed by estar + gerund (e.g. Ela está dançando ‘She is dancing’). There are also phonological differences such as the patterns of vowel reduction in unstressed syllables, with the overall tendency for European Portuguese to reduce and even delete vowels where Brazilian Portuguese keeps them. So for the purposes of this thought experiment, let’s consider them different languages, albeit closely related ones. Our imaginary Martian would take note of the differences between the two kinds of Portuguese, as well as of the extensive similarities between them and would deduce—correctly, of course!—that both European and Brazilian Portuguese descend from a common ancestor. The question for our Martian is this: where was the linguistic homeland of this Proto-Portuguese? There are three theories that the Martian might reasonably advance: (a) that the homeland was in Europe, specifically in the western part of the Iberian Peninsula, where European Portuguese is spoken today, from which it would follow that it was brought over to Brazil by some later colonizers; (b) that the homeland was in Brazil, which would entail that European Portuguese emerged upon the arrival of Brazilian migrants to Europe; or (c) that the homeland was in some other location altogether, meaning that it was brought over to both the Iberian Peninsula and Brazil and did not survive in its original homeland. (The latter model has been proposed for the Austronesian languages, which are thought to have originated in coastal mainland China, although no Austronesian languages survive there.) In the absence of some strong evidence for theory (c), the Martian would probably decide in favor of either (a) or (b) as less convoluted hypotheses; and the fact that there are far more speakers of Portuguese in Brazil than in Europe might sway him in favor of theory (b). Of course, unlike the hypothetical Martian, we know from the historical record that theory (a) is the correct one.
Note, however, that it is impossible for the Proto-Portuguese homeland to be in both Europe and Brazil: separated by such a formidable geographical obstacle as the Atlantic Ocean, how would the speakers make and keep their language the same for any length of time? Because of the unrelenting linguistic change, whenever persistent and intensive contact between groups is not possible, their ways of speaking necessarily diverge. It should also be noted here that geographical separation is not the only factor that drives language diversification. Far more important is the factor of time. However geographically contiguous a linguistic territory is, “evolution will not be uniform throughout the territory but will vary from zone to zone,” Ferdinand de Saussure wrote in his Course in General Linguistics, “no records indicate that any language has ever changed in the same way throughout its territory.” (p. 199) A great example of such diversification in situ is the emergence of contemporary Irish Gaelic dialects in Ireland, described in the work of Jim McCloskey.
Going back to our original PIE homeland problem, the two main competitors for its location—the Pontic Steppes area in southern Russia and central Anatolia in what is now Turkey—may not seem as far apart as Portugal and Brazil, but several millennia ago the Black Sea that separates them presented as formidable a barrier to human travel as the Atlantic did in later times (see the work of Kate Davison and her colleagues: Davison et al. 2006). Therefore, it is not possible that PIE was spoken in the Steppes area and in Anatolia and did not diverge. Moreover, the two main competing theories place PIE not only in different locations but also at different times, differing by about 3,000 years. It is not possible for PIE to be spoken in Anatolia and then in exactly the same way in the Steppes area three millennia later (Anthony 2007 makes a similar argument).
One other point is crucial to stress: whatever solution one provides for any linguistic homeland problem, PIE homeland included, it does not deny that there were population movements before and after the postulated homeland—but these migration patterns, interesting though they may be, are irrelevant to the homeland problem. Going back to our Portuguese thought experiment, the fact that the speakers of European Portuguese came to the Iberian Peninsula from elsewhere in Europe some centuries earlier is no more relevant to the Proto-Portuguese homeland problem than the migration of their earlier ancestors from the Steppes 5,500 years ago (or Anatolia 8,500 years ago if you side with that theory), or the Out-of-Africa exodus of homo sapiens some 70,000 years ago. Nor are later migrations, such as the arrival of African slaves to Brazil, relevant to the Proto-Portuguese homeland problem. (Of course, the story of African immigrants who learned Portuguese as their second-language is very important for understanding some of the changes that occurred in Brazilian Portuguese, as compared to its European variant, but that is a different issue.)
As for the PIE debate, it is plausible that present-day Europeans descend from successive ways of migrants from different places: the pre-Neolithic inhabitants, as well as early Neolithic agriculturalists from Anatolia and later Neolithic arrivals from the Steppes—as genetic studies seem to indicate. However, only one of these groups could be the “vectors for the spread of Indo-European languages into Europe” (Haak et al. 2015), by definition.
Anthony, David W. (2007) The Horse, the Wheel, and Language: How Bronze-Age Riders from the Eurasian Steppes Shaped the Modern World. Princeton, NJ: Princeton University Press.
Davison, Kate; Pavel Dolukhanov, Graeme R. Sarson, and Anvar Shukurov (2006) The role of waterways in the spread of the Neolithic. Journal of Archaeological Science 33: 641-652.
Haak, Wolfgang et al. (2015) Massive migration from the steppe is a source for Indo-European languages in Europe. Biorxiv Online.
Piazza, Alberto and Luigi Luca Cavalli-Sforza (2006) Diffusion of genes and languages in human evolution. Proceedings of the 6th International Conference on the Evolution of Language. Pp. 255-266.
Saussure, Ferdinand de (1959) Course in General Linguistics. New York: Philosophical Library.
« Je Suis “Random Folk in a Deli”—What...
Rethinking the Role of Broca’s Area—Or... »