The date of PIE.

From the literature, there seems to be a consensus that it is not possible to place PIE unity before 4,000 BCE. The most commonly cited evidence lies in the words for wheels and wheeled vehicles, which (1) archaeology forbids quite nearly anyone anywhere before this time, and (2) that these words in no way give any evidence of having been passed from one IE language to another but are inherited directly by each IE language from the common ancestor.

Another oft-cited piece of evidence is that PIE apparently had only one word for metal (undoubtedly copper) and that the breakup of PIE occured before additional metal words appear in the daughter language (a distinct word for gold seems to have individually developed in the daughter languages before copper and bronze are clearly distinguished).

This places enormous contraints on when (and where) you can place the final unity of PIE. The standard model is the late Eneolithic Ponto-Caspian steppe ca 3500 BCE, with the Yamnaya culture being named as the prime candidate.

If a language is to remain unified, i.e., remain a single language, its speakers must be in proximity to each other. The amount of territory any language can permanently occupy is related to mobility. The more mobile a language's speakers are, the more territory that language can occupy. We know the Indo-Iranians were extremely mobile, and the vast territory their language historically occupied reflects this. Mobility is technological (wheeled vehicles, draft animals, means of war), geographical (difficult terrain has always functioned as a natural border) and politics (already occupied lands can successfully resist expansionist intrusions unless your war-making means are better than theirs).

The final question is the time depth necessary for separate language branches (at the level of Germanic or Slavic) to develop. This is highly controversial. Most commentators dismiss 'glottochronology' on methodological grounds. All we have to go on is the depth of changes we see in historically attested languages, viz, the emergence of Romance out of Latin, Slavic from proto-Slavic, the Indic group out of Sanskrit, etc.

One the one hand, we see the long period of time in which these branches have maintained their integrity, and not broken into separate branches; there are mutually unintelligable languages within each individual branche (some of them extraordinarly mutually unintelligable), but nary a one has reached the point where we are ready to say it's a founder of a separate branch of the Indo-European family (Digression: are their any candidate languages for such status?).

On the other hand, we also see the extraordinarily conservative effect writing can have on languages. Romance has a 2,000 old Latin adstrate, while Hindu has a similar relationship with Sanskrit.

A third phenomenon is the founder effect, where once a language or one branch of a language family establishes itself in a region, and once the population of native-speakers reaches a 'critical mass', the possibility for one language permanently replacing another is greatly inhibited.

PIE of course had its own mother language. I've read suggestions that sometime around the 5500 BCE level, we might look for unity with the so-called Tyrrhenian group (Etruscan, Lemnian, maybe Pelasgian), and perhaps, the final break with Uralic. I've not read too much on this (just comments from those who have read of it), but the date proposed was arrived at before knowledge of the Black Sea flood (and its 5500 BCE radiocarbon-attested date) was attained. This event undoubtedly had an effect on the coalescence of PIE, though the magnitude, small or large, remains to be investigated.

Mallory questions the IEness of the Corded Ware horizon, albeit inexplicitly. This is not to say this horizon did not speak a language related to IE, only that no traces of these langauges remain (except perhaps as the substratum in Germanic).

What follows here are my own speculations.

The PIE speakers need not be the authors of the Yamnaya culture, but rather, only its inheritors. By modern standards -- and even by ancient standards -- this part of the world was very thinly populated, and until the development of steppe nomadism, virtually terra nullis (unpopulated territory claimed by no one). Minor changes and small movements could have a very large effect.

My knowledge of what these changes were is weak, but I think it has a lot to do with the perfection of wheeled vehicles, the technology for harnessing animals to such vehicles, combined with an advanced understanding of raising herds of domestic animals.

At a certain point, this technology reached takeoff. They were better-fed than their neighbors and consequently enjoyed a population explosion. At the same time, they were masters of the technology that let them move out from the original center very rapidly, colonizing previously empty zones, while co-opting other groups within their extended material culture. The pattern is at once one of founders occupying hitherto vacant lands, and one of elite dominance, probably combined with a modest but real superiority of numbers.

I admit that the differences in the economies of the northern forest, as well as what happens west of the Carpathians needs further explanation. One thought I have is that their very success on the steppe started the historic pattern for the steppe -- groups being pushed westward by disturbances to the east. If we accept that Germanic represents a group that replaced their original language (probably via elite dominance) at a relatively late date, the only other problem is Anatolic -- a problem that perhaps solves itself by having Anatolic being south of the Danube, separated from the remaining body of IE speakers at an earlier date (say, 3500 BCE).

As I said, this is mostly personal speculation. At the very least, I'm just thinking out loud. I'd like to think I'm conversing, having 'bull sessions', where one talks out what one is thinking about. This is the only way you reach really valid conclusions.

Mark Odegard.