Daniel Dennett on memes and our mediated realities, in his recent book Breaking the Spell: Religion as a Natural Phenomenon:
…it seems best to include all these replicators [computer virii and online scams/social engineering] under the rubric of memes, noting that some of them make only indirect use of human vectors, and hence are only indirectly elements of human culture. We are beginning to see this porous boundary crossed in the other direction as well: it used to be true that the differential replication of such classic memes as songs, poems and recipes depended on their winning the competition for residence in human brains, but now that a multitude of search engines on the Web have interposed themselves between authors and their (human) audiences, competing with one another for reputation as high-quality sources of cultural items, significant fitness differences between memes can accumulate independently of any human appreciation or cognizance at all. The day may soon come when a cleverly turned phrase in a book gets indexed by many search engines, and thereupon enters the language as a new cliché, without anybody human having read the book.
[My emphasis]
Indeed. And more generally, the many weirdnesses of words, halt-footed bearers of heavy memes that they are, when processed through dumb tech. From some of our recent work, it’s become very clear that issues around search engines and memes will loom large into our near future — one of the challenges for anyone trying to track the spread and evolution of (human-generated) memes online is the attempt to identify the current lexical correlates of a particular meme, and to understand how and when that lexical structure changes over time as the meme itself mutates. It’s a hard problem — something like building a parse tree for a Gerald Manley Hopkins poem, maybe (recommended reading: Lexical Ambiguity in Poetry). Only over time. Harder.
And for those of us in the communications business — in the absence of real automated semantic analysis — there’s the challenge of trying to instrument our messages with trackable terms or phrases which survive intact as the messages themselves spread and mutate in the (human and digital) wilds.
Early days. Having just read both Dennett and Hamlet’s Mill on holiday, it’s pretty clear that for true memetic (not simply lexical) invariance over time, we need to engineer myth. Then build search engines which interrogate mythic structures.
It’s all getting a bit Snow Crash. Whatever. Count me in.