I want to clarify something about memes, as I have referred to them in two posts in a row. Much of the criticism of the idea stems from the fact that they aren't "real": Language changes, technologies change, and each time someone uses a word or builds a prison or pours whisky into a tumbler, it is subtly different (in some way) to every previous time. So "memes" aren't stable things you can point to, and therefore it is unreasonable to speak of an Indo-European expansion if only "memes" expanded. "Memes" aren't real.
Well, I agree with most of those criticisms, because they're true, if arcane. Every time anyone does anything, it's based on a unique calculation or set of calculations occurring in their brain and in others' that has never previously occurred, and which is susceptible to an enormous range of influences. So memes aren't metaphysically real; I don't think anyone has claimed that they are, and in fact it's because of arguments like those above that meme theory has changed and been explained by philosophers like Dennett.
These criticisms also apply to a population, and that's part of why Darwinism is so tricky for people; we're used to thinking of "species" and "populations" as being discrete entities with properties of their own, but of course, they're not. They're abstractions, with members of the species/population having unique genomes and other properties. Even if Indo-European languages were moved across the place by a group of people with ancestry going back entirely to the speakers of Proto-Indo-European, they were still different groups moving into different areas. PIE speakers had sex and procreated, and each time you do that you create a new human being with a different genome to both of its parents. Even a population in situ with no new populations coming in will be a different thing after a generation, with a different set of genomes and experiences to everyone who lived before. The current population of Nauru is different to the population of Micronesian sailors who first settled it, and that would be the case even if no one else had ever visited the islands since the first sailors landed.
So populations aren't "real" either, and yet most people have no problem with saying something like, "Spanish people moved to Mexico after Cortes' conquest" or "the native people of Tahiti are Polynesians who sailed across the Pacific to populate the island from Samoa". Memes are almost as real as populations, I'd say, and we can talk about languages and technologies about as easily as we can talk about populations, or about tables, or rock formations, or any other things that reduce in all of their properties to their parts. Languages, memes, populations - these are abstractions that we use to think with, and that correspond with what they attempt to describe in a meaningful but not an absolute fashion.
It is only by bearing this in mind that we can understand things like the founder effect, which applies about as well to language, "memes", etc, as to population genetics. This is because people never fully understand everything the people in their community do, and they never take in everything equally, just as the members of a population don't have identical genomes. When a small number of them leave and go somewhere else, they take with them only what they remember or what they like - whatever is stored in their brains.