Memetics and Probablities

As a musician, writer, academic, whatever-the-hell-I-decide-I-am-that-day, the biggest burning question to me is how exactly do ideas generate and propogate? I've read a buttload of psychology (the discipline of which is ill-equipped at the moment to deal with this), history, philosophy, semiotics and the like, and not come across anyone who's really started to tackle this question. It's a thorny issue in general, when you have to start with "well what constitutes an idea?" and a whole other related definitive questions, but still, it remains. 

More interesting still is when people converge on an idea in entirely different circumstances on limited amounts of the same information. This has occurred time and time again throughout the histories of art, music, science, mathematics... So many bizarre incidences of people arriving at the same solutions to issues entirely independently, and with little recourse to what we'd consider linear logic from those initial questions. 

Eventually, when I found my way to memetics, I thought that this could potentially give at least a narrative solution to this idea. The solution could be seen as that similar local ecologies of memes (the local memepool) and a similar environment could yield similar mutations - a process in evolutionary biology called convergent evolution. 

As Steven Jan (eminent musical memeticist) notes in his paper "Replicating Sonorities" coequality of memes is rather difficult to distinguish, and therefore probabilistic means shall have to be implied when distinguishing between the equalness of memes at different levels of meaning. However, this idea of probability in memetic creation and transmission is actually a really elegant solution to the above problems. 

For example, is there a formula for how far a memeplex (a collection of memes/ideas that replicate and adapt together - think "genre") has to come before a measured evolution and stable mutation is likely and suggested? How do ideas that are "in the air" in similar cultures with similar inputs occur independently? There are numerous notable cases where this happens (especially in the sciences - see Kuhn and scientific revolutions etc), and it's certainly advantageous to understand this mechanism for it happens often enough that there surely is one... HOW DOES THIS HAPPEN?

Probability actually really fucking important - how probable is it that a meme will copied? What conditions need to be satisfied on a gross basis? (individual differences occur, but in talking of memetics and musical receptivity altogether we're in the realm of abstracts already, so how abstractly likely is it for any given meme to be copied at all, and how is this affected/how does it affect variability?)

Apologies for the echo-chamber post, but sometimes I think it's interesting to try follow someone else's thoughts into abstract realms just to sort of make your brain reach into places it's not accustomed to.