I got quite excited when I first found out about it, because it was a very good match for the way I was envisioning the development of traditional formal features in the early Greek epic tradition at the time. I had been reading a bit too much Gregory Nagy -- at the time he had the habit of claiming that the Homeric epics were formed more by tradition than by poets -- and after I had monologued on the subject in late-night conversations, some friends got me reading Daniel Dennett.

But my excitement wore off once I realised that it was just an alternate way of talking about the development of ideas, and that beyond the different perspective -- the "meme's eye perspective" -- it wasn't actually bringing anything new to the table.

Methodologically, the real problem with memes is that there's no parsimony. With genes, or other forms of reproduction, parsimony is king: it's the linchpin that makes phylogenies possible. You use parsimony to analyse similarities and difference between a set of tokens (or species, in the case of biology), and with the assumption that "fewer differences = closer phylogenetic relationship", you use the magic of collation or principal component analysis or whatever to construct a phylogeny.

But that parsimony principle -- the "fewer differences = closer phylogenetic relationship" principle, the idea that reproduction is conservative -- just doesn't apply nearly as strongly with memes. That's one of the problems I had with a study in comparative mythology that I talked about in a thread last week. Here's a passage in Homer that my supervisor pointed out to me as a caution, when I had enthused to her about memes (Odyssey 1.351-2):

For people praise the song more when it is the newest one to make the rounds for its listeners.

The idea of memetics relies on the principle that reproduction is conservative. But in reality, in the realm of ideas to some extent and certainly in the realm art, innovation and novelty are prized. Sure there's conservatism too, but it's notoriously difficult to interpret. You can't take type scenes in Homer -- setpiece scenes that follow a regular pattern of motifs -- and construct a phylogeny out of them, on the understanding that they represent a self-propagating meme in the epic tradition, because poets are always doing something new with them. A musicologist that I know has the same problem with formalist memes: he insists that there's no such thing as sonata form. I mean, there is, but at the same time there isn't: Haydn didn't have an authoritative rule book on how that meme ought to look, and Homer had no rule book on how similes and type scenes ought to look. Formalism has its place, but you need to be cautious as all hell with it: linguistic ideas like emergent grammar (grammaticalisation) are more appropriate than rigid phylogenies.

Anyway, "meme" is really just an alternate name for something that people were already talking about, so I'm glad the word has lost its pseudo-scientific pretensions now.